Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -39,30 +39,23 @@ try:
|
|
| 39 |
except Exception: whisper_stt_client = None
|
| 40 |
|
| 41 |
# --- Helper Functions ---
|
| 42 |
-
def generate_story_from_llm(name, grade_str, topic
|
| 43 |
-
progress(0.0, desc="Starting story creation...")
|
| 44 |
default_passage_val = ""
|
| 45 |
-
default_audio_gen_update = gr.update(interactive=False, visible=False)
|
| 46 |
-
default_audio_player_update = gr.update(value=None, visible=False)
|
| 47 |
if not LLM_API_CONFIGURED:
|
| 48 |
-
|
| 49 |
-
return "LLM API key not configured...", default_audio_gen_update, default_audio_player_update
|
| 50 |
try:
|
| 51 |
if grade_str.startswith("Grade "):
|
| 52 |
grade = int(grade_str.replace("Grade ", ""))
|
| 53 |
else:
|
| 54 |
grade = int(grade_str)
|
| 55 |
except ValueError:
|
| 56 |
-
|
| 57 |
-
return "Invalid grade level selected.", default_audio_gen_update, default_audio_player_update
|
| 58 |
if grade <= 2: word_target, max_llm_tokens = "around 40-60 words", 100
|
| 59 |
elif grade <= 5: word_target, max_llm_tokens = "around 80-100 words", 200
|
| 60 |
elif grade <= 8: word_target, max_llm_tokens = "around 100-120 words", 250
|
| 61 |
else: word_target, max_llm_tokens = "around 120-150 words", 300
|
| 62 |
|
| 63 |
-
progress(0.1, desc="Setting up AI story generator...")
|
| 64 |
story_text_result = default_passage_val
|
| 65 |
-
audio_gen_btn_update = default_audio_gen_update
|
| 66 |
try:
|
| 67 |
model = genai.GenerativeModel(model_name="gemini-1.5-flash-latest")
|
| 68 |
prompt = (
|
|
@@ -78,9 +71,7 @@ def generate_story_from_llm(name, grade_str, topic, progress=gr.Progress(track_t
|
|
| 78 |
"HARM_CATEGORY_DANGEROUS_CONTENT"
|
| 79 |
]]
|
| 80 |
generation_config = genai.types.GenerationConfig(candidate_count=1, max_output_tokens=max_llm_tokens, temperature=0.7)
|
| 81 |
-
progress(0.3, desc="AI is writing your story...")
|
| 82 |
response = model.generate_content(prompt, generation_config=generation_config, safety_settings=safety_settings)
|
| 83 |
-
progress(0.8, desc="Polishing your story...")
|
| 84 |
if response.candidates and response.candidates[0].content.parts:
|
| 85 |
story = response.text
|
| 86 |
if response.prompt_feedback and response.prompt_feedback.block_reason:
|
|
@@ -89,51 +80,40 @@ def generate_story_from_llm(name, grade_str, topic, progress=gr.Progress(track_t
|
|
| 89 |
story_text_result = f"The LLM couldn't generate a story for '{topic}'. Try another topic or rephrase. ✨"
|
| 90 |
else:
|
| 91 |
story_text_result = story.strip()
|
| 92 |
-
audio_gen_btn_update = gr.update(interactive=True, visible=True)
|
| 93 |
else:
|
| 94 |
if response.prompt_feedback and response.prompt_feedback.block_reason:
|
| 95 |
story_text_result = f"Story idea for '{topic}' got blocked (Reason: {response.prompt_feedback.block_reason}). Try a different topic. 😊"
|
| 96 |
else:
|
| 97 |
story_text_result = "Hmm, LLM had trouble with that topic. Maybe try another one? 🤔"
|
| 98 |
-
|
| 99 |
-
return story_text_result, audio_gen_btn_update, default_audio_player_update
|
| 100 |
except Exception as e:
|
| 101 |
-
|
| 102 |
-
return f"Oh no! 😟 Error generating story. Details: {e}", default_audio_gen_update, default_audio_player_update
|
| 103 |
|
| 104 |
-
def
|
|
|
|
| 105 |
global tts_client
|
| 106 |
-
progress(0.0, desc="🔊 Preparing voice synthesis...")
|
| 107 |
|
| 108 |
-
if not
|
| 109 |
-
progress(1.0, desc="Complete")
|
| 110 |
return None
|
| 111 |
|
| 112 |
-
progress(0.1, desc="🔊 Initializing audio generation...")
|
| 113 |
-
|
| 114 |
# Reconnect to TTS client if needed
|
| 115 |
if not tts_client:
|
| 116 |
-
progress(0.2, desc="🔗 Connecting to advanced voice service...")
|
| 117 |
try:
|
| 118 |
tts_client = Client("NihalGazi/Text-To-Speech-Unlimited")
|
| 119 |
-
progress(0.3, desc="🔗 Connected to voice service...")
|
| 120 |
except Exception as e:
|
| 121 |
print(f"Failed to connect to TTS service: {e}")
|
| 122 |
-
progress(1.0, desc="Complete")
|
| 123 |
return None
|
| 124 |
-
|
| 125 |
if not tts_client:
|
| 126 |
-
progress(1.0, desc="Complete")
|
| 127 |
return None
|
| 128 |
|
| 129 |
-
progress(0.4, desc="🎙️ AI is reading your story aloud...")
|
| 130 |
try:
|
| 131 |
# Try the correct API configuration with emotion parameter
|
| 132 |
api_methods = [
|
| 133 |
-
{"params": [
|
| 134 |
-
{"params": [
|
| 135 |
-
{"params": [
|
| 136 |
-
{"params": [
|
| 137 |
]
|
| 138 |
|
| 139 |
audio_filepath = None
|
|
@@ -163,8 +143,6 @@ def text_to_speech_using_space(text_to_speak, progress=gr.Progress(track_tqdm=Tr
|
|
| 163 |
continue
|
| 164 |
|
| 165 |
if audio_filepath:
|
| 166 |
-
progress(0.9, desc="🎵 Voice generation complete!")
|
| 167 |
-
progress(1.0, desc="🔊 Audio ready!")
|
| 168 |
print(f"FINAL: Returning audio file path: {audio_filepath}")
|
| 169 |
return audio_filepath
|
| 170 |
else:
|
|
@@ -175,13 +153,11 @@ def text_to_speech_using_space(text_to_speak, progress=gr.Progress(track_tqdm=Tr
|
|
| 175 |
print(f"TTS error: {e}")
|
| 176 |
# Try to reconnect on error
|
| 177 |
try:
|
| 178 |
-
progress(0.6, desc="🔄 Reconnecting to voice service...")
|
| 179 |
tts_client = Client("NihalGazi/Text-To-Speech-Unlimited")
|
| 180 |
if tts_client:
|
| 181 |
-
progress(0.8, desc="🎙️ Retrying voice generation...")
|
| 182 |
# Try the most basic approach with emotion parameter
|
| 183 |
audio_result = tts_client.predict(
|
| 184 |
-
|
| 185 |
"alloy", # voice
|
| 186 |
"neutral", # emotion
|
| 187 |
api_name="/text_to_speech_app"
|
|
@@ -197,7 +173,6 @@ def text_to_speech_using_space(text_to_speak, progress=gr.Progress(track_tqdm=Tr
|
|
| 197 |
audio_filepath = audio_result[0]
|
| 198 |
|
| 199 |
if audio_filepath:
|
| 200 |
-
progress(1.0, desc="🔊 Audio ready!")
|
| 201 |
print(f"RETRY SUCCESS: Returning audio file path: {audio_filepath}")
|
| 202 |
return audio_filepath
|
| 203 |
|
|
@@ -205,23 +180,17 @@ def text_to_speech_using_space(text_to_speak, progress=gr.Progress(track_tqdm=Tr
|
|
| 205 |
print(f"TTS retry failed: {retry_error}")
|
| 206 |
pass
|
| 207 |
|
| 208 |
-
progress(1.0, desc="Audio generation failed")
|
| 209 |
print("TTS failed completely - returning None")
|
| 210 |
return None
|
| 211 |
|
| 212 |
-
def speech_to_text_whisper_space(audio_filepath,
|
| 213 |
-
progress(0.1, desc="Sending your reading for transcription...")
|
| 214 |
if not whisper_stt_client:
|
| 215 |
-
progress(1.0, desc="Complete")
|
| 216 |
return "Speech-to-text service is not available. 🛠️"
|
| 217 |
if not audio_filepath:
|
| 218 |
-
progress(1.0, desc="Complete")
|
| 219 |
return "No recording received for transcription. 🎤"
|
| 220 |
for attempt in range(max_retries):
|
| 221 |
try:
|
| 222 |
-
progress(0.2 + (attempt * 0.1), desc=f"Transcribing your voice (Whisper) - Attempt {attempt + 1}...")
|
| 223 |
result = whisper_stt_client.predict(audio_filepath, api_name="/predict")
|
| 224 |
-
progress(0.9, desc="Transcription complete.")
|
| 225 |
if isinstance(result, tuple) and len(result) > 0:
|
| 226 |
transcribed_text = result[0] if result[0] else ""
|
| 227 |
elif isinstance(result, list) and len(result) > 0:
|
|
@@ -229,13 +198,10 @@ def speech_to_text_whisper_space(audio_filepath, progress=gr.Progress(track_tqdm
|
|
| 229 |
elif isinstance(result, str):
|
| 230 |
transcribed_text = result
|
| 231 |
else:
|
| 232 |
-
progress(1.0, desc="Complete")
|
| 233 |
return "Hmm, STT service returned unexpected format. 🤔"
|
| 234 |
-
progress(1.0, desc="Transcription complete!")
|
| 235 |
return transcribed_text if transcribed_text else "No speech detected in the recording. 🤫"
|
| 236 |
except Exception:
|
| 237 |
continue
|
| 238 |
-
progress(1.0, desc="Complete")
|
| 239 |
return "Unexpected error during transcription. Please try again! 🔄"
|
| 240 |
|
| 241 |
def clean_text_for_comparison(text):
|
|
@@ -246,25 +212,64 @@ def clean_text_for_comparison(text):
|
|
| 246 |
def compare_texts_for_feedback(original_text, student_text):
|
| 247 |
original_words, student_words = clean_text_for_comparison(original_text), clean_text_for_comparison(student_text)
|
| 248 |
if not student_words: return "It sounds like you didn't record or it was very quiet! 🤫 Try recording again nice and clear!", ""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 249 |
matcher = difflib.SequenceMatcher(None, original_words, student_words, autojunk=False)
|
| 250 |
-
|
| 251 |
-
|
| 252 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 253 |
pronunciation_guide = {
|
| 254 |
'the': 'thuh or thee', 'through': 'threw', 'though': 'thoh', 'thought': 'thawt',
|
| 255 |
'knight': 'night', 'know': 'noh', 'write': 'right', 'wrong': 'rawng', 'what': 'wot',
|
| 256 |
'where': 'wair', 'when': 'wen', 'why': 'wy', 'who': 'hoo', 'laugh': 'laff',
|
| 257 |
'enough': 'ee-nuff', 'cough': 'koff', 'rough': 'ruff', 'tough': 'tuff', 'magic': 'maj-ik',
|
| 258 |
'school': 'skool', 'friend': 'frend', 'said': 'sed', 'says': 'sez', 'once': 'wunts',
|
| 259 |
-
'was': 'wuz', 'were': 'wur', 'you': 'yoo', 'your': 'yor', 'there': 'thair', 'their': 'thair', 'they': 'thay'
|
|
|
|
|
|
|
|
|
|
| 260 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 261 |
def get_pronunciation_tip(word):
|
| 262 |
word_lower = word.lower()
|
| 263 |
if word_lower in pronunciation_guide:
|
| 264 |
-
return
|
| 265 |
elif len(word) > 6:
|
| 266 |
-
|
| 267 |
vowels = 'aeiou'
|
|
|
|
| 268 |
current_syllable = ''
|
| 269 |
for i, char in enumerate(word_lower):
|
| 270 |
current_syllable += char
|
|
@@ -273,83 +278,264 @@ def compare_texts_for_feedback(original_text, student_text):
|
|
| 273 |
syllables.append(current_syllable)
|
| 274 |
current_syllable = ''
|
| 275 |
if current_syllable: syllables.append(current_syllable)
|
| 276 |
-
if len(syllables) > 1: return
|
| 277 |
-
|
| 278 |
-
|
| 279 |
-
|
| 280 |
-
elif word_lower.startswith('wr'): return "🗣️ In 'wr-' words, the 'w' is silent"
|
| 281 |
-
elif word_lower.startswith('kn'): return "🗣️ In 'kn-' words, the 'k' is silent"
|
| 282 |
-
return f"🎯 Focus on each sound in \"{word}\""
|
| 283 |
for tag, i1, i2, j1, j2 in matcher.get_opcodes():
|
| 284 |
orig_seg_words, stud_seg_words = original_words[i1:i2], student_words[j1:j2]
|
| 285 |
orig_seg_text, stud_seg_text = " ".join(orig_seg_words), " ".join(stud_seg_words)
|
|
|
|
| 286 |
if tag == 'equal':
|
| 287 |
-
|
| 288 |
-
|
| 289 |
-
|
| 290 |
-
|
| 291 |
-
|
| 292 |
-
|
| 293 |
-
|
| 294 |
-
|
| 295 |
-
|
| 296 |
-
|
| 297 |
-
|
| 298 |
-
|
| 299 |
-
|
| 300 |
-
|
| 301 |
-
|
| 302 |
-
|
| 303 |
-
|
| 304 |
-
|
| 305 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 306 |
final_text = " ".join(highlighted_parts)
|
| 307 |
-
|
| 308 |
-
|
| 309 |
-
|
| 310 |
-
|
| 311 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 312 |
"""
|
| 313 |
return feedback_html, final_text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 314 |
else:
|
| 315 |
-
|
| 316 |
-
|
| 317 |
-
|
| 318 |
-
|
| 319 |
-
"",
|
| 320 |
-
"🔍 **What to work on:**"
|
| 321 |
]
|
| 322 |
-
for line in feedback_lines: feedback_parts.append(f"• {line}")
|
| 323 |
-
if pronunciation_tips:
|
| 324 |
-
feedback_parts.extend([
|
| 325 |
-
"",
|
| 326 |
-
"🎤 **Pronunciation Helper**",
|
| 327 |
-
"Here's how to say the tricky words:"
|
| 328 |
-
])
|
| 329 |
-
for tip in pronunciation_tips[:5]: feedback_parts.append(f"• {tip}")
|
| 330 |
-
feedback_parts.extend([
|
| 331 |
-
"",
|
| 332 |
-
"💡 **Pro tip:** Listen to the story audio again and pay special attention to these words!"
|
| 333 |
-
])
|
| 334 |
-
feedback_parts.extend([
|
| 335 |
-
"",
|
| 336 |
-
"🎯 **Practice Suggestions**",
|
| 337 |
-
"• 🎧 Listen to the AI reading first",
|
| 338 |
-
"• 🔤 Practice saying difficult words slowly",
|
| 339 |
-
"• 📖 Read the story again at your own pace",
|
| 340 |
-
"• 🔄 Try recording again when you're ready!"
|
| 341 |
-
])
|
| 342 |
-
feedback_html = "\n".join(feedback_parts)
|
| 343 |
-
return feedback_html, final_text
|
| 344 |
|
| 345 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 346 |
if not student_audio_path: return "🎤 Please record your reading first!", ""
|
| 347 |
if not original_passage_state: return "Hmm, the original story is missing. 😟 Please generate a story first.", ""
|
| 348 |
-
transcribed_text = speech_to_text_whisper_space(student_audio_path
|
| 349 |
stt_errors = ["couldn't understand", "had trouble", "service isn't working", "service is busy", "didn't get any recording", "filepath type issue"]
|
| 350 |
if any(err in (transcribed_text or "").lower() for err in stt_errors): return transcribed_text, ""
|
| 351 |
-
|
| 352 |
-
|
| 353 |
|
| 354 |
css = """
|
| 355 |
body, .gradio-container {
|
|
@@ -357,17 +543,54 @@ body, .gradio-container {
|
|
| 357 |
font-family: -apple-system, BlinkMacSystemFont, 'San Francisco', 'Segoe UI', 'Roboto', Arial, sans-serif !important;
|
| 358 |
}
|
| 359 |
.main-header {
|
| 360 |
-
background:
|
| 361 |
-
border-radius:
|
| 362 |
-
box-shadow: 0 8px 32px 0 rgba(
|
| 363 |
-
padding:
|
| 364 |
-
margin
|
|
|
|
| 365 |
text-align: center;
|
| 366 |
border: none !important;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 367 |
}
|
| 368 |
-
.main-header h1 {font-size: 2.2rem !important; font-weight: 700 !important; color: #23232b !important;}
|
| 369 |
-
.main-header p {color: #6b7280 !important; font-size: 1.08rem !important; margin-bottom: 8px !important;}
|
| 370 |
-
.tech-badge {background: #e0e7ef !important; color: #4f8fff !important; border-radius: 12px !important; padding: 4px 12px !important; font-size: 12px !important; font-weight: 600 !important;}
|
| 371 |
.gr-block, .gr-panel {background: white !important; border-radius: 18px !important; box-shadow: 0 2px 8px 0 rgba(60,60,90,0.07) !important; border: none !important; padding: 28px 22px !important;}
|
| 372 |
.section-header {background: transparent !important; border: none !important; padding: 0 !important; margin-bottom: 16px !important;}
|
| 373 |
.section-header h3 {color: #1e293b !important; font-size: 1.14rem !important; font-weight: 600 !important;}
|
|
@@ -404,7 +627,7 @@ body, .gradio-container {
|
|
| 404 |
background: linear-gradient(90deg, #e0e7ef, #dde5f2) !important;
|
| 405 |
color: #2a3140 !important;
|
| 406 |
transition: all 0.15s cubic-bezier(0.4,0.0,0.2,1) !important;
|
| 407 |
-
transform: translateY(0) !important
|
| 408 |
}
|
| 409 |
|
| 410 |
.gr-button[variant="secondary"]:hover {
|
|
@@ -480,7 +703,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 480 |
audio_out = gr.Audio(label="🎵 Story Audio", type="filepath", visible=True, autoplay=False)
|
| 481 |
gr.Markdown("""
|
| 482 |
<div style="margin: 20px 0 0 0; padding: 10px 20px; background: #f4f7fa; border-radius: 16px;">
|
| 483 |
-
<b>➡️ Next:</b> Record yourself reading below for
|
| 484 |
</div>
|
| 485 |
""")
|
| 486 |
stud_audio_in = gr.Audio(sources=["microphone"], type="filepath", label="🎤 Your Recording")
|
|
@@ -489,53 +712,56 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 489 |
assess_btn = gr.Button("🔍 Analyze Reading", variant="primary", size="lg", interactive=False)
|
| 490 |
recording_status = gr.Markdown("", elem_id="recording_status")
|
| 491 |
analysis_status = gr.Markdown("", elem_id="analysis_status")
|
| 492 |
-
with gr.TabItem("📊 Analysis & Feedback", elem_id="analysis_tab"):
|
| 493 |
-
gr.Markdown("""
|
| 494 |
-
<div class="section-header">
|
| 495 |
-
<h3>📈 Analysis</h3>
|
| 496 |
-
<p>See your performance and areas to improve</p>
|
| 497 |
-
</div>
|
| 498 |
-
""")
|
| 499 |
-
feedback_out = gr.Markdown(
|
| 500 |
-
value="""
|
| 501 |
-
<div style="text-align: center; color: #6b7280;">
|
| 502 |
-
<h4>Analysis Results</h4>
|
| 503 |
-
<p>Your feedback will appear here.</p>
|
| 504 |
-
</div>
|
| 505 |
-
""",
|
| 506 |
-
elem_id="feedback_output"
|
| 507 |
-
)
|
| 508 |
-
highlighted_out = gr.Markdown(
|
| 509 |
-
value="""
|
| 510 |
-
<div style="text-align: center; color: #6b7280;">
|
| 511 |
-
<h4>Word-by-Word Analysis</h4>
|
| 512 |
-
<p>Get color-coded feedback below.</p>
|
| 513 |
-
</div>
|
| 514 |
-
""",
|
| 515 |
-
elem_id="highlighted_passage_output"
|
| 516 |
-
)
|
| 517 |
-
gr.Markdown("""
|
| 518 |
-
<div style="background: #f7fafc; border-radius: 16px; padding: 16px 12px 10px 12px; margin: 22px 0 18px 0; box-shadow: 0 2px 6px 0 rgba(60,60,90,0.04);">
|
| 519 |
-
<b>Color code:</b>
|
| 520 |
-
<div style="display: flex; gap: 14px; flex-wrap: wrap; margin-top: 8px;">
|
| 521 |
-
<span style="background: #90EE90; padding: 5px 14px; border-radius: 12px; font-size: 13px; font-weight: 500; color: #155724;">Perfect Match</span>
|
| 522 |
-
<span style="background: #FFE4B5; padding: 5px 14px; border-radius: 12px; font-size: 13px; font-weight: 500; color: #856404;">Substitution</span>
|
| 523 |
-
<span style="background: #FFA0B4; padding: 5px 14px; border-radius: 12px; font-size: 13px; font-weight: 500; color: #721c24;">Skipped Word</span>
|
| 524 |
-
<span style="background: #DDA0DD; padding: 5px 14px; border-radius: 12px; font-size: 13px; font-weight: 500; color: #5f006a;">Extra Word</span>
|
| 525 |
-
</div>
|
| 526 |
-
</div>
|
| 527 |
-
""")
|
| 528 |
-
gr.Markdown("""
|
| 529 |
-
<div style="margin: 14px 0; padding: 14px 22px; background: #f8fafc; border-radius: 14px;">
|
| 530 |
-
<span style="color: #0a58ca; font-weight: 500;">Goals:</span>
|
| 531 |
-
<ul style="margin: 7px 0 0 18px; color: #6b7280;">
|
| 532 |
-
<li>Word accuracy above 90%</li>
|
| 533 |
-
<li>Speak clearly and with confidence</li>
|
| 534 |
-
<li>Practice as much as you like</li>
|
| 535 |
-
</ul>
|
| 536 |
-
</div>
|
| 537 |
-
""")
|
| 538 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 539 |
with gr.TabItem("ℹ️ About & How It Works", elem_id="about_tab"):
|
| 540 |
gr.Markdown("""
|
| 541 |
<div class="section-header">
|
|
@@ -544,23 +770,10 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 544 |
</div>
|
| 545 |
""")
|
| 546 |
|
| 547 |
-
gr.Markdown("""
|
| 548 |
-
## 🎯 What This Platform Does
|
| 549 |
-
|
| 550 |
-
ReadRight is an AI-powered tool designed to help students improve their reading skills through:
|
| 551 |
-
|
| 552 |
-
- **✨ Personalized Story Generation**: Creates age-appropriate reading passages tailored to your grade level and interests
|
| 553 |
-
- **🔊 Audio Pronunciation Models**: Provides clear audio examples of proper pronunciation
|
| 554 |
-
- **⚡ Real-time Speech Analysis**: Analyzes your reading accuracy and identifies areas for improvement
|
| 555 |
-
- **🎯 Detailed Feedback**: Offers specific pronunciation tips and practice suggestions
|
| 556 |
-
|
| 557 |
-
## 🏗️ Reading Practice Application Workflow
|
| 558 |
-
""")
|
| 559 |
-
|
| 560 |
-
# Use HTML component for the SVG
|
| 561 |
gr.HTML("""
|
| 562 |
-
<div style="
|
| 563 |
-
<
|
|
|
|
| 564 |
<!-- Background -->
|
| 565 |
<rect width="1400" height="600" fill="#fafafa"/>
|
| 566 |
|
|
@@ -656,6 +869,15 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 656 |
""")
|
| 657 |
|
| 658 |
gr.Markdown("""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 659 |
---
|
| 660 |
|
| 661 |
## 🔧 Key Components
|
|
@@ -685,7 +907,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 685 |
passage_state = story_text
|
| 686 |
return story_text, audio_btn_update, audio_player_update, passage_state
|
| 687 |
|
| 688 |
-
def assess_reading_with_analysis(original_passage_state, student_audio_path
|
| 689 |
if not student_audio_path:
|
| 690 |
return (
|
| 691 |
"""
|
|
@@ -707,12 +929,8 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 707 |
""
|
| 708 |
)
|
| 709 |
|
| 710 |
-
# Immediate feedback that analysis is starting
|
| 711 |
-
progress(0.05, desc="Analysis starting...")
|
| 712 |
-
|
| 713 |
# Start transcription
|
| 714 |
-
|
| 715 |
-
transcribed_text = speech_to_text_whisper_space(student_audio_path, progress=progress)
|
| 716 |
|
| 717 |
stt_errors = ["couldn't understand", "had trouble", "service isn't working", "service is busy", "didn't get any recording", "filepath type issue"]
|
| 718 |
if any(err in (transcribed_text or "").lower() for err in stt_errors):
|
|
@@ -727,9 +945,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 727 |
""
|
| 728 |
)
|
| 729 |
|
| 730 |
-
progress(0.6, desc="Analyzing your reading accuracy...")
|
| 731 |
feedback, highlighted_passage = compare_texts_for_feedback(original_passage_state, transcribed_text)
|
| 732 |
-
progress(1.0, desc="Assessment complete!")
|
| 733 |
|
| 734 |
analysis_msg = """
|
| 735 |
<div class="status-indicator status-success">
|
|
@@ -739,6 +955,28 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 739 |
"""
|
| 740 |
return (analysis_msg, feedback, highlighted_passage)
|
| 741 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 742 |
def update_recording_status(audio_file):
|
| 743 |
if audio_file is not None:
|
| 744 |
return (
|
|
@@ -810,89 +1048,29 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css, title="ReadRight") as app:
|
|
| 810 |
gr.update(visible=False),
|
| 811 |
gr.update(interactive=False)
|
| 812 |
)
|
| 813 |
-
|
| 814 |
-
def handle_audio_generation(story_text, progress=gr.Progress(track_tqdm=True)):
|
| 815 |
-
"""Handle audio generation with visual progress indicator"""
|
| 816 |
-
if not story_text or not story_text.strip():
|
| 817 |
-
return (
|
| 818 |
-
gr.update(value=None, visible=True),
|
| 819 |
-
gr.update(value="", visible=False)
|
| 820 |
-
)
|
| 821 |
-
|
| 822 |
-
# Generate the audio file
|
| 823 |
-
audio_filepath = text_to_speech_using_space(story_text, progress)
|
| 824 |
-
print(f"AUDIO HANDLER: Received audio file path: {audio_filepath}")
|
| 825 |
-
|
| 826 |
-
if audio_filepath:
|
| 827 |
-
print(f"AUDIO HANDLER: Updating audio component with file: {audio_filepath}")
|
| 828 |
-
success_msg = """
|
| 829 |
-
<div style="background: #f0fdf4; border: 1px solid #22c55e; border-radius: 12px; padding: 12px 20px; margin: 8px 0; text-align: center;">
|
| 830 |
-
<span style="color: #15803d; font-weight: 500;">✅ Audio ready! You can now listen to your story.</span>
|
| 831 |
-
</div>
|
| 832 |
-
"""
|
| 833 |
-
return (
|
| 834 |
-
gr.update(value=audio_filepath, visible=True),
|
| 835 |
-
gr.update(value=success_msg, visible=True)
|
| 836 |
-
)
|
| 837 |
-
else:
|
| 838 |
-
print("AUDIO HANDLER: No audio file received, returning None")
|
| 839 |
-
error_msg = """
|
| 840 |
-
<div style="background: #fef2f2; border: 1px solid #ef4444; border-radius: 12px; padding: 12px 20px; margin: 8px 0; text-align: center;">
|
| 841 |
-
<span style="color: #dc2626; font-weight: 500;">❌ Audio generation failed. Please try again.</span>
|
| 842 |
-
</div>
|
| 843 |
-
"""
|
| 844 |
-
return (
|
| 845 |
-
gr.update(value=None, visible=True),
|
| 846 |
-
gr.update(value=error_msg, visible=True)
|
| 847 |
-
)
|
| 848 |
|
| 849 |
-
|
| 850 |
-
"""Generate story and automatically create audio in one seamless flow"""
|
| 851 |
-
progress(0.0, desc="Starting story creation...")
|
| 852 |
-
|
| 853 |
-
# First generate the story using the existing function
|
| 854 |
-
story_result = generate_story_from_llm(name, grade, topic, progress)
|
| 855 |
-
if not story_result:
|
| 856 |
-
return "", gr.update(value=None, visible=True), ""
|
| 857 |
-
|
| 858 |
-
# Extract story text from the result tuple
|
| 859 |
-
story_text = story_result[0] if isinstance(story_result, tuple) else story_result
|
| 860 |
-
|
| 861 |
-
# Check if story generation was successful
|
| 862 |
-
if not story_text or any(err in story_text.lower() for err in ["error", "blocked", "couldn't", "api key not configured"]):
|
| 863 |
-
return story_text, gr.update(value=None, visible=True), ""
|
| 864 |
-
|
| 865 |
-
# Story generated successfully, now automatically generate audio
|
| 866 |
-
progress(0.5, desc="Story complete! Now generating audio...")
|
| 867 |
-
|
| 868 |
-
try:
|
| 869 |
-
# Generate audio automatically
|
| 870 |
-
audio_filepath = text_to_speech_using_space(story_text, progress)
|
| 871 |
-
|
| 872 |
-
if audio_filepath:
|
| 873 |
-
print(f"AUTO AUDIO: Successfully generated audio: {audio_filepath}")
|
| 874 |
-
return story_text, gr.update(value=audio_filepath, visible=True), story_text
|
| 875 |
-
else:
|
| 876 |
-
print("AUTO AUDIO: Audio generation failed, but story is still available")
|
| 877 |
-
return story_text, gr.update(value=None, visible=True), story_text
|
| 878 |
-
|
| 879 |
-
except Exception as e:
|
| 880 |
-
print(f"AUTO AUDIO ERROR: {e}")
|
| 881 |
-
return story_text, gr.update(value=None, visible=True), story_text
|
| 882 |
-
|
| 883 |
-
# Event handlers with automatic audio generation
|
| 884 |
gen_btn.click(
|
| 885 |
-
fn=
|
| 886 |
inputs=[s_name, s_grade, s_topic],
|
| 887 |
-
outputs=[passage_out
|
| 888 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 889 |
)
|
| 890 |
|
| 891 |
assess_btn.click(
|
| 892 |
fn=assess_reading_with_analysis,
|
| 893 |
inputs=[original_passage_state, stud_audio_in],
|
| 894 |
-
outputs=[analysis_status, feedback_out, highlighted_out]
|
| 895 |
-
show_progress=True
|
| 896 |
)
|
| 897 |
|
| 898 |
stud_audio_in.change(
|
|
|
|
| 39 |
except Exception: whisper_stt_client = None
|
| 40 |
|
| 41 |
# --- Helper Functions ---
|
| 42 |
+
def generate_story_from_llm(name, grade_str, topic):
|
|
|
|
| 43 |
default_passage_val = ""
|
|
|
|
|
|
|
| 44 |
if not LLM_API_CONFIGURED:
|
| 45 |
+
return "LLM API key not configured..."
|
|
|
|
| 46 |
try:
|
| 47 |
if grade_str.startswith("Grade "):
|
| 48 |
grade = int(grade_str.replace("Grade ", ""))
|
| 49 |
else:
|
| 50 |
grade = int(grade_str)
|
| 51 |
except ValueError:
|
| 52 |
+
return "Invalid grade level selected."
|
|
|
|
| 53 |
if grade <= 2: word_target, max_llm_tokens = "around 40-60 words", 100
|
| 54 |
elif grade <= 5: word_target, max_llm_tokens = "around 80-100 words", 200
|
| 55 |
elif grade <= 8: word_target, max_llm_tokens = "around 100-120 words", 250
|
| 56 |
else: word_target, max_llm_tokens = "around 120-150 words", 300
|
| 57 |
|
|
|
|
| 58 |
story_text_result = default_passage_val
|
|
|
|
| 59 |
try:
|
| 60 |
model = genai.GenerativeModel(model_name="gemini-1.5-flash-latest")
|
| 61 |
prompt = (
|
|
|
|
| 71 |
"HARM_CATEGORY_DANGEROUS_CONTENT"
|
| 72 |
]]
|
| 73 |
generation_config = genai.types.GenerationConfig(candidate_count=1, max_output_tokens=max_llm_tokens, temperature=0.7)
|
|
|
|
| 74 |
response = model.generate_content(prompt, generation_config=generation_config, safety_settings=safety_settings)
|
|
|
|
| 75 |
if response.candidates and response.candidates[0].content.parts:
|
| 76 |
story = response.text
|
| 77 |
if response.prompt_feedback and response.prompt_feedback.block_reason:
|
|
|
|
| 80 |
story_text_result = f"The LLM couldn't generate a story for '{topic}'. Try another topic or rephrase. ✨"
|
| 81 |
else:
|
| 82 |
story_text_result = story.strip()
|
|
|
|
| 83 |
else:
|
| 84 |
if response.prompt_feedback and response.prompt_feedback.block_reason:
|
| 85 |
story_text_result = f"Story idea for '{topic}' got blocked (Reason: {response.prompt_feedback.block_reason}). Try a different topic. 😊"
|
| 86 |
else:
|
| 87 |
story_text_result = "Hmm, LLM had trouble with that topic. Maybe try another one? 🤔"
|
| 88 |
+
return story_text_result
|
|
|
|
| 89 |
except Exception as e:
|
| 90 |
+
return f"Oh no! 😟 Error generating story. Details: {e}"
|
|
|
|
| 91 |
|
| 92 |
+
def text_to_speech_using_space_simple(text):
|
| 93 |
+
"""Simplified TTS function - Gradio will show its default loading indicator"""
|
| 94 |
global tts_client
|
|
|
|
| 95 |
|
| 96 |
+
if not text or not text.strip():
|
|
|
|
| 97 |
return None
|
| 98 |
|
|
|
|
|
|
|
| 99 |
# Reconnect to TTS client if needed
|
| 100 |
if not tts_client:
|
|
|
|
| 101 |
try:
|
| 102 |
tts_client = Client("NihalGazi/Text-To-Speech-Unlimited")
|
|
|
|
| 103 |
except Exception as e:
|
| 104 |
print(f"Failed to connect to TTS service: {e}")
|
|
|
|
| 105 |
return None
|
| 106 |
+
|
| 107 |
if not tts_client:
|
|
|
|
| 108 |
return None
|
| 109 |
|
|
|
|
| 110 |
try:
|
| 111 |
# Try the correct API configuration with emotion parameter
|
| 112 |
api_methods = [
|
| 113 |
+
{"params": [text, "alloy", "happy"], "api_name": "/text_to_speech_app"},
|
| 114 |
+
{"params": [text, "alloy", "neutral"], "api_name": "/text_to_speech_app"},
|
| 115 |
+
{"params": [text, "nova", "neutral"], "api_name": "/text_to_speech_app"},
|
| 116 |
+
{"params": [text], "api_name": "/predict"}
|
| 117 |
]
|
| 118 |
|
| 119 |
audio_filepath = None
|
|
|
|
| 143 |
continue
|
| 144 |
|
| 145 |
if audio_filepath:
|
|
|
|
|
|
|
| 146 |
print(f"FINAL: Returning audio file path: {audio_filepath}")
|
| 147 |
return audio_filepath
|
| 148 |
else:
|
|
|
|
| 153 |
print(f"TTS error: {e}")
|
| 154 |
# Try to reconnect on error
|
| 155 |
try:
|
|
|
|
| 156 |
tts_client = Client("NihalGazi/Text-To-Speech-Unlimited")
|
| 157 |
if tts_client:
|
|
|
|
| 158 |
# Try the most basic approach with emotion parameter
|
| 159 |
audio_result = tts_client.predict(
|
| 160 |
+
text,
|
| 161 |
"alloy", # voice
|
| 162 |
"neutral", # emotion
|
| 163 |
api_name="/text_to_speech_app"
|
|
|
|
| 173 |
audio_filepath = audio_result[0]
|
| 174 |
|
| 175 |
if audio_filepath:
|
|
|
|
| 176 |
print(f"RETRY SUCCESS: Returning audio file path: {audio_filepath}")
|
| 177 |
return audio_filepath
|
| 178 |
|
|
|
|
| 180 |
print(f"TTS retry failed: {retry_error}")
|
| 181 |
pass
|
| 182 |
|
|
|
|
| 183 |
print("TTS failed completely - returning None")
|
| 184 |
return None
|
| 185 |
|
| 186 |
+
def speech_to_text_whisper_space(audio_filepath, max_retries=3):
|
|
|
|
| 187 |
if not whisper_stt_client:
|
|
|
|
| 188 |
return "Speech-to-text service is not available. 🛠️"
|
| 189 |
if not audio_filepath:
|
|
|
|
| 190 |
return "No recording received for transcription. 🎤"
|
| 191 |
for attempt in range(max_retries):
|
| 192 |
try:
|
|
|
|
| 193 |
result = whisper_stt_client.predict(audio_filepath, api_name="/predict")
|
|
|
|
| 194 |
if isinstance(result, tuple) and len(result) > 0:
|
| 195 |
transcribed_text = result[0] if result[0] else ""
|
| 196 |
elif isinstance(result, list) and len(result) > 0:
|
|
|
|
| 198 |
elif isinstance(result, str):
|
| 199 |
transcribed_text = result
|
| 200 |
else:
|
|
|
|
| 201 |
return "Hmm, STT service returned unexpected format. 🤔"
|
|
|
|
| 202 |
return transcribed_text if transcribed_text else "No speech detected in the recording. 🤫"
|
| 203 |
except Exception:
|
| 204 |
continue
|
|
|
|
| 205 |
return "Unexpected error during transcription. Please try again! 🔄"
|
| 206 |
|
| 207 |
def clean_text_for_comparison(text):
|
|
|
|
| 212 |
def compare_texts_for_feedback(original_text, student_text):
|
| 213 |
original_words, student_words = clean_text_for_comparison(original_text), clean_text_for_comparison(student_text)
|
| 214 |
if not student_words: return "It sounds like you didn't record or it was very quiet! 🤫 Try recording again nice and clear!", ""
|
| 215 |
+
|
| 216 |
+
# Enhanced analysis metrics
|
| 217 |
+
total_original_words = len(original_words)
|
| 218 |
+
total_student_words = len(student_words)
|
| 219 |
+
|
| 220 |
matcher = difflib.SequenceMatcher(None, original_words, student_words, autojunk=False)
|
| 221 |
+
highlighted_parts = []
|
| 222 |
+
|
| 223 |
+
# Detailed tracking
|
| 224 |
+
correct_words = 0
|
| 225 |
+
substituted_words = 0
|
| 226 |
+
missed_words = 0
|
| 227 |
+
extra_words = 0
|
| 228 |
+
|
| 229 |
+
# New improved tracking
|
| 230 |
+
challenging_words = []
|
| 231 |
+
skill_areas = {
|
| 232 |
+
'accuracy': {'score': 0, 'tips': []},
|
| 233 |
+
'fluency': {'score': 0, 'tips': []},
|
| 234 |
+
'pronunciation': {'score': 0, 'tips': []}
|
| 235 |
+
}
|
| 236 |
+
|
| 237 |
+
# Enhanced pronunciation guide with more words
|
| 238 |
pronunciation_guide = {
|
| 239 |
'the': 'thuh or thee', 'through': 'threw', 'though': 'thoh', 'thought': 'thawt',
|
| 240 |
'knight': 'night', 'know': 'noh', 'write': 'right', 'wrong': 'rawng', 'what': 'wot',
|
| 241 |
'where': 'wair', 'when': 'wen', 'why': 'wy', 'who': 'hoo', 'laugh': 'laff',
|
| 242 |
'enough': 'ee-nuff', 'cough': 'koff', 'rough': 'ruff', 'tough': 'tuff', 'magic': 'maj-ik',
|
| 243 |
'school': 'skool', 'friend': 'frend', 'said': 'sed', 'says': 'sez', 'once': 'wunts',
|
| 244 |
+
'was': 'wuz', 'were': 'wur', 'you': 'yoo', 'your': 'yor', 'there': 'thair', 'their': 'thair', 'they': 'thay',
|
| 245 |
+
'because': 'bee-koz', 'beautiful': 'byoo-ti-ful', 'different': 'dif-er-ent', 'important': 'im-por-tant',
|
| 246 |
+
'people': 'pee-pul', 'together': 'too-geth-er', 'water': 'waw-ter', 'favorite': 'fay-vor-it',
|
| 247 |
+
'journey': 'jur-nee', 'treasure': 'trezh-er', 'adventure': 'ad-ven-cher', 'mysterious': 'mis-teer-ee-us'
|
| 248 |
}
|
| 249 |
+
|
| 250 |
+
def identify_challenging_words(text_words):
|
| 251 |
+
"""Identify potentially difficult words from the story for proactive help"""
|
| 252 |
+
challenging = []
|
| 253 |
+
for word in text_words:
|
| 254 |
+
word_lower = word.lower()
|
| 255 |
+
# Add words that are commonly mispronounced or complex
|
| 256 |
+
if (len(word) > 6 or # Long words
|
| 257 |
+
word_lower in pronunciation_guide or # Known difficult words
|
| 258 |
+
'tion' in word_lower or 'ough' in word_lower or # Tricky endings
|
| 259 |
+
word_lower.startswith(('wr', 'kn', 'ph')) or # Silent letters
|
| 260 |
+
'gh' in word_lower or 'th' in word_lower): # Difficult sounds
|
| 261 |
+
if word_lower not in challenging:
|
| 262 |
+
challenging.append(word_lower)
|
| 263 |
+
return challenging[:5] # Limit to 5 most relevant words
|
| 264 |
+
|
| 265 |
def get_pronunciation_tip(word):
|
| 266 |
word_lower = word.lower()
|
| 267 |
if word_lower in pronunciation_guide:
|
| 268 |
+
return pronunciation_guide[word_lower]
|
| 269 |
elif len(word) > 6:
|
| 270 |
+
# Simple syllable breakdown
|
| 271 |
vowels = 'aeiou'
|
| 272 |
+
syllables = []
|
| 273 |
current_syllable = ''
|
| 274 |
for i, char in enumerate(word_lower):
|
| 275 |
current_syllable += char
|
|
|
|
| 278 |
syllables.append(current_syllable)
|
| 279 |
current_syllable = ''
|
| 280 |
if current_syllable: syllables.append(current_syllable)
|
| 281 |
+
if len(syllables) > 1: return '-'.join(syllables)
|
| 282 |
+
return word_lower
|
| 283 |
+
|
| 284 |
+
# Process each operation in the diff for highlighting
|
|
|
|
|
|
|
|
|
|
| 285 |
for tag, i1, i2, j1, j2 in matcher.get_opcodes():
|
| 286 |
orig_seg_words, stud_seg_words = original_words[i1:i2], student_words[j1:j2]
|
| 287 |
orig_seg_text, stud_seg_text = " ".join(orig_seg_words), " ".join(stud_seg_words)
|
| 288 |
+
|
| 289 |
if tag == 'equal':
|
| 290 |
+
correct_words += len(orig_seg_words)
|
| 291 |
+
highlighted_parts.append(f'<span style="background: #22c55e; color: white; padding: 3px 6px; border-radius: 6px; margin: 2px; font-weight: 500;">{orig_seg_text}</span>')
|
| 292 |
+
elif tag == 'replace':
|
| 293 |
+
substituted_words += len(orig_seg_words)
|
| 294 |
+
highlighted_parts.append(f'<span style="background: #f59e0b; color: white; padding: 3px 6px; border-radius: 6px; margin: 2px; font-weight: 500; text-decoration: line-through;">{orig_seg_text}</span> <span style="background: #ef4444; color: white; padding: 3px 6px; border-radius: 6px; margin: 2px; font-weight: 500;">→{stud_seg_text}</span>')
|
| 295 |
+
elif tag == 'delete':
|
| 296 |
+
missed_words += len(orig_seg_words)
|
| 297 |
+
highlighted_parts.append(f'<span style="background: #ef4444; color: white; padding: 3px 6px; border-radius: 6px; margin: 2px; font-weight: 500; text-decoration: line-through;">{orig_seg_text}</span> <span style="font-style: italic; color: #9ca3af; font-size: 0.9em;">(skipped)</span>')
|
| 298 |
+
elif tag == 'insert':
|
| 299 |
+
extra_words += len(stud_seg_words)
|
| 300 |
+
highlighted_parts.append(f'<span style="background: #8b5cf6; color: white; padding: 3px 6px; border-radius: 6px; margin: 2px; font-weight: 500; font-style: italic;">+{stud_seg_text}</span>')
|
| 301 |
+
|
| 302 |
+
# Calculate comprehensive metrics
|
| 303 |
+
accuracy_percentage = round((correct_words / total_original_words) * 100, 1) if total_original_words > 0 else 0
|
| 304 |
+
|
| 305 |
+
# Determine performance level
|
| 306 |
+
if accuracy_percentage >= 95:
|
| 307 |
+
performance_level = "🏆 Excellent"
|
| 308 |
+
performance_color = "#10b981"
|
| 309 |
+
performance_message = "Outstanding reading! You're reading like a champion!"
|
| 310 |
+
elif accuracy_percentage >= 85:
|
| 311 |
+
performance_level = "🌟 Very Good"
|
| 312 |
+
performance_color = "#3b82f6"
|
| 313 |
+
performance_message = "Great job! You're doing really well with your reading."
|
| 314 |
+
elif accuracy_percentage >= 70:
|
| 315 |
+
performance_level = "💪 Good Progress"
|
| 316 |
+
performance_color = "#f59e0b"
|
| 317 |
+
performance_message = "Nice work! Keep practicing to improve even more."
|
| 318 |
+
elif accuracy_percentage >= 50:
|
| 319 |
+
performance_level = "📚 Keep Practicing"
|
| 320 |
+
performance_color = "#ef4444"
|
| 321 |
+
performance_message = "You're learning! More practice will help you improve."
|
| 322 |
+
else:
|
| 323 |
+
performance_level = "🚀 Just Getting Started"
|
| 324 |
+
performance_color = "#8b5cf6"
|
| 325 |
+
performance_message = "Every reader starts somewhere! Keep trying and you'll get better."
|
| 326 |
+
|
| 327 |
+
# Generate challenging words for proactive help
|
| 328 |
+
challenging_words = identify_challenging_words(original_words)
|
| 329 |
+
|
| 330 |
+
# Assess skill areas
|
| 331 |
+
skill_areas['accuracy']['score'] = accuracy_percentage
|
| 332 |
+
if accuracy_percentage < 90:
|
| 333 |
+
skill_areas['accuracy']['tips'] = ['Practice reading slowly and clearly', 'Follow along with the text while listening']
|
| 334 |
+
|
| 335 |
+
skill_areas['fluency']['score'] = max(0, 100 - (missed_words * 10))
|
| 336 |
+
if missed_words > 2:
|
| 337 |
+
skill_areas['fluency']['tips'] = ['Try reading the story multiple times', 'Practice difficult words separately first']
|
| 338 |
+
|
| 339 |
+
skill_areas['pronunciation']['score'] = max(0, 100 - (substituted_words * 15))
|
| 340 |
+
if substituted_words > 1:
|
| 341 |
+
skill_areas['pronunciation']['tips'] = ['Listen carefully to each word sound', 'Break long words into smaller parts']
|
| 342 |
+
|
| 343 |
final_text = " ".join(highlighted_parts)
|
| 344 |
+
|
| 345 |
+
# Perfect reading case
|
| 346 |
+
if accuracy_percentage == 100:
|
| 347 |
+
feedback_html = f"""
|
| 348 |
+
<div style="background: linear-gradient(135deg, #10b981, #059669); padding: 24px; border-radius: 16px; color: white; text-align: center; margin-bottom: 20px;">
|
| 349 |
+
<h2 style="margin: 0 0 8px 0; font-size: 1.8rem;">🎉 PERFECT READING! 🎉</h2>
|
| 350 |
+
<p style="margin: 0; font-size: 1.1rem; opacity: 0.9;">Amazing! You read every single word correctly!</p>
|
| 351 |
+
</div>
|
| 352 |
+
|
| 353 |
+
<div style="background: #f0fdf4; border: 2px solid #22c55e; border-radius: 12px; padding: 20px; margin-bottom: 16px;">
|
| 354 |
+
<h3 style="color: #15803d; margin: 0 0 12px 0;">📊 Your Reading Score</h3>
|
| 355 |
+
<div style="display: grid; grid-template-columns: 1fr 1fr; gap: 16px; margin-bottom: 16px;">
|
| 356 |
+
<div style="text-align: center;">
|
| 357 |
+
<div style="font-size: 2rem; font-weight: bold; color: #15803d;">100%</div>
|
| 358 |
+
<div style="font-size: 0.9rem; color: #166534;">Word Accuracy</div>
|
| 359 |
+
</div>
|
| 360 |
+
<div style="text-align: center;">
|
| 361 |
+
<div style="font-size: 2rem; font-weight: bold, color: #15803d;">{total_original_words}/{total_original_words}</div>
|
| 362 |
+
<div style="font-size: 0.9rem, color: #166534;">Words Correct</div>
|
| 363 |
+
</div>
|
| 364 |
+
</div>
|
| 365 |
+
<div style="text-align: center; padding: 12px; background: #dcfce7; border-radius: 8px;">
|
| 366 |
+
<strong style="color: #15803d;">🏆 Reading Champion Level!</strong>
|
| 367 |
+
</div>
|
| 368 |
+
</div>
|
| 369 |
+
|
| 370 |
+
<div style="background: #fffbeb; border-radius: 12px; padding: 16px;">
|
| 371 |
+
<h4 style="color: #92400e; margin: 0 0 8px 0;">🎯 What's Next?</h4>
|
| 372 |
+
<ul style="margin: 8px 0; padding-left: 20px; color: #78350f;">
|
| 373 |
+
<li>Try a more challenging story topic</li>
|
| 374 |
+
<li>Practice reading faster while staying accurate</li>
|
| 375 |
+
<li>Help a friend or family member practice reading</li>
|
| 376 |
+
<li>Celebrate your excellent reading skills! 🎊</li>
|
| 377 |
+
</ul>
|
| 378 |
+
</div>
|
| 379 |
"""
|
| 380 |
return feedback_html, final_text
|
| 381 |
+
|
| 382 |
+
# Improved analysis with non-repetitive, skill-focused feedback
|
| 383 |
+
feedback_html = f"""
|
| 384 |
+
<div style="background: linear-gradient(135deg, {performance_color}, {performance_color}dd); padding: 20px; border-radius: 16px; color: white; text-align: center; margin-bottom: 20px;">
|
| 385 |
+
<h2 style="margin: 0 0 8px 0; font-size: 1.6rem;">{performance_level}</h2>
|
| 386 |
+
<p style="margin: 0; font-size: 1rem; opacity: 0.9;">{performance_message}</p>
|
| 387 |
+
</div>
|
| 388 |
+
|
| 389 |
+
<div style="background: #f8fafc; border-radius: 12px; padding: 20px; margin-bottom: 20px;">
|
| 390 |
+
<h3 style="color: #1e293b; margin: 0 0 16px 0;">📊 Reading Dashboard</h3>
|
| 391 |
+
<div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(120px, 1fr)); gap: 16px; margin-bottom: 16px;">
|
| 392 |
+
<div style="text-align: center; background: white; padding: 12px; border-radius: 8px; box-shadow: 0 1px 3px rgba(0,0,0,0.1);">
|
| 393 |
+
<div style="font-size: 1.8rem; font-weight: bold; color: {performance_color};">{accuracy_percentage}%</div>
|
| 394 |
+
<div style="font-size: 0.85rem; color: #64748b;">Accuracy</div>
|
| 395 |
+
</div>
|
| 396 |
+
<div style="text-align: center; background: white; padding: 12px; border-radius: 8px; box-shadow: 0 1px 3px rgba(0,0,0,0.1);">
|
| 397 |
+
<div style="font-size: 1.8rem; font-weight: bold; color: #22c55e;">{correct_words}</div>
|
| 398 |
+
<div style="font-size: 0.85rem; color: #64748b;">Words Correct</div>
|
| 399 |
+
</div>
|
| 400 |
+
<div style="text-align: center; background: white; padding: 12px; border-radius: 8px; box-shadow: 0 1px 3px rgba(0,0,0,0.1);">
|
| 401 |
+
<div style="font-size: 1.8rem; font-weight: bold; color: #ef4444;">{missed_words}</div>
|
| 402 |
+
<div style="font-size: 0.85rem; color: #64748b;">Missed</div>
|
| 403 |
+
</div>
|
| 404 |
+
<div style="text-align: center; background: white; padding: 12px; border-radius: 8px; box-shadow: 0 1px 3px rgba(0,0,0,0.1);">
|
| 405 |
+
<div style="font-size: 1.8rem; font-weight: bold; color: #f59e0b;">{substituted_words}</div>
|
| 406 |
+
<div style="font-size: 0.85rem; color: #64748b;">Changed</div>
|
| 407 |
+
</div>
|
| 408 |
+
</div>
|
| 409 |
+
|
| 410 |
+
<!-- Color Legend -->
|
| 411 |
+
<div style="background: #ffffff; border-radius: 8px; padding: 12px; margin-top: 16px;">
|
| 412 |
+
<h4 style="color: #374151; margin: 0 0 8px 0; font-size: 0.9rem;">📖 Word Color Guide:</h4>
|
| 413 |
+
<div style="display: flex; flex-wrap: wrap; gap: 8px; font-size: 0.8rem;">
|
| 414 |
+
<span style="background: #22c55e; color: white; padding: 2px 8px; border-radius: 4px;">✓ Correct</span>
|
| 415 |
+
<span style="background: #ef4444; color: white; padding: 2px 8px; border-radius: 4px;">✗ Missed</span>
|
| 416 |
+
<span style="background: #f59e0b; color: white; padding: 2px 8px; border-radius: 4px;">~ Changed</span>
|
| 417 |
+
<span style="background: #8b5cf6; color: white; padding: 2px 8px; border-radius: 4px;">+ Added</span>
|
| 418 |
+
</div>
|
| 419 |
+
</div>
|
| 420 |
+
</div>
|
| 421 |
+
"""
|
| 422 |
+
|
| 423 |
+
# Smart Focus Areas - skill-based instead of error repetition
|
| 424 |
+
improvement_areas = []
|
| 425 |
+
if accuracy_percentage < 85:
|
| 426 |
+
improvement_areas.append("🎯 **Reading Accuracy**: Focus on reading each word carefully")
|
| 427 |
+
if missed_words > 2:
|
| 428 |
+
improvement_areas.append("📖 **Reading Fluency**: Practice reading without skipping words")
|
| 429 |
+
if substituted_words > 2:
|
| 430 |
+
improvement_areas.append("🗣️ **Pronunciation**: Work on saying words clearly")
|
| 431 |
+
if extra_words > 1:
|
| 432 |
+
improvement_areas.append("👁️ **Focus & Attention**: Follow the text closely while reading")
|
| 433 |
+
|
| 434 |
+
if improvement_areas:
|
| 435 |
+
feedback_html += f"""
|
| 436 |
+
<div style="background: #fef9c3; border-left: 4px solid #eab308; padding: 16px; border-radius: 8px; margin-bottom: 16px;">
|
| 437 |
+
<h4 style="color: #a16207; margin: 0 0 12px 0;">🎯 Smart Focus Areas</h4>
|
| 438 |
+
<div style="color: #a16207;">
|
| 439 |
+
"""
|
| 440 |
+
for area in improvement_areas[:3]: # Limit to 3 most important areas
|
| 441 |
+
feedback_html += f" • {area}<br>"
|
| 442 |
+
|
| 443 |
+
feedback_html += """
|
| 444 |
+
</div>
|
| 445 |
+
</div>
|
| 446 |
+
"""
|
| 447 |
+
|
| 448 |
+
# Proactive Pronunciation Helper - based on story words, not just errors
|
| 449 |
+
if challenging_words:
|
| 450 |
+
feedback_html += f"""
|
| 451 |
+
<div style="background: #e0f2fe; border-radius: 12px; padding: 16px; margin-bottom: 16px;">
|
| 452 |
+
<h4 style="color: #0277bd; margin: 0 0 12px 0;">🗣️ Story Word Pronunciation Guide</h4>
|
| 453 |
+
<p style="color: #0277bd; font-size: 0.9rem; margin: 0 0 12px 0;">Here are some words from your story that might be tricky:</p>
|
| 454 |
+
<div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(200px, 1fr)); gap: 12px;">
|
| 455 |
+
"""
|
| 456 |
+
for word in challenging_words:
|
| 457 |
+
pronunciation = get_pronunciation_tip(word)
|
| 458 |
+
feedback_html += f"""
|
| 459 |
+
<div style="background: #b3e5fc; padding: 8px 12px; border-radius: 6px;">
|
| 460 |
+
<strong style="color: #01579b;">{word.upper()}</strong><br>
|
| 461 |
+
<span style="color: #0277bd; font-size: 0.85rem;">"say: {pronunciation}"</span>
|
| 462 |
+
</div>"""
|
| 463 |
+
|
| 464 |
+
feedback_html += """
|
| 465 |
+
</div>
|
| 466 |
+
<div style="margin-top: 12px; padding: 8px; background: #b3e5fc; border-radius: 6px; font-size: 0.9rem;">
|
| 467 |
+
💡 <strong>Practice tip:</strong> Listen to the AI reading these words and repeat them slowly!
|
| 468 |
+
</div>
|
| 469 |
+
</div>
|
| 470 |
+
"""
|
| 471 |
+
|
| 472 |
+
# Progress Insights for Parents
|
| 473 |
+
feedback_html += f"""
|
| 474 |
+
<div style="background: #f0f9ff; border-radius: 12px; padding: 16px; margin-bottom: 16px;">
|
| 475 |
+
<h4 style="color: #0369a1; margin: 0 0 12px 0;">📈 Reading Skills Progress</h4>
|
| 476 |
+
<div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(150px, 1fr)); gap: 12px;">
|
| 477 |
+
<div style="background: white; padding: 10px; border-radius: 6px; text-align: center;">
|
| 478 |
+
<div style="font-size: 1.2rem; font-weight: bold; color: #0369a1;">{skill_areas['accuracy']['score']:.0f}%</div>
|
| 479 |
+
<div style="font-size: 0.8rem; color: #64748b;">Word Accuracy</div>
|
| 480 |
+
</div>
|
| 481 |
+
<div style="background: white; padding: 10px; border-radius: 6px; text-align: center;">
|
| 482 |
+
<div style="font-size: 1.2rem; font-weight: bold; color: #0369a1;">{skill_areas['fluency']['score']:.0f}%</div>
|
| 483 |
+
<div style="font-size: 0.8rem; color: #64748b;">Reading Flow</div>
|
| 484 |
+
</div>
|
| 485 |
+
<div style="background: white; padding: 10px; border-radius: 6px; text-align: center;">
|
| 486 |
+
<div style="font-size: 1.2rem; font-weight: bold; color: #0369a1;">{skill_areas['pronunciation']['score']:.0f}%</div>
|
| 487 |
+
<div style="font-size: 0.8rem; color: #64748b;">Pronunciation</div>
|
| 488 |
+
</div>
|
| 489 |
+
</div>
|
| 490 |
+
</div>
|
| 491 |
+
"""
|
| 492 |
+
|
| 493 |
+
# Personalized Next Steps
|
| 494 |
+
if accuracy_percentage >= 85:
|
| 495 |
+
next_steps = [
|
| 496 |
+
"🎧 Practice reading along with the audio for better timing",
|
| 497 |
+
"📚 Try a slightly more challenging story topic",
|
| 498 |
+
"🗣️ Focus on reading with expression and emotion"
|
| 499 |
+
]
|
| 500 |
+
elif accuracy_percentage >= 70:
|
| 501 |
+
next_steps = [
|
| 502 |
+
"🎧 Listen to the AI reading first, then read yourself",
|
| 503 |
+
"🔤 Practice the tricky words from above separately",
|
| 504 |
+
"📱 Record yourself multiple times and compare"
|
| 505 |
+
]
|
| 506 |
else:
|
| 507 |
+
next_steps = [
|
| 508 |
+
"🎧 Listen to the audio several times before recording",
|
| 509 |
+
"👁️ Follow along with the text while listening",
|
| 510 |
+
"⏰ Take your time - read slowly and clearly"
|
|
|
|
|
|
|
| 511 |
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 512 |
|
| 513 |
+
feedback_html += f"""
|
| 514 |
+
<div style="background: #f0f9ff; border-radius: 12px; padding: 16px;">
|
| 515 |
+
<h4 style="color: #0369a1; margin: 0 0 12px 0;">🎮 Your Reading Quest - Next Steps!</h4>
|
| 516 |
+
<div style="color: #0369a1;">
|
| 517 |
+
"""
|
| 518 |
+
for step in next_steps:
|
| 519 |
+
feedback_html += f" • {step}<br>"
|
| 520 |
+
|
| 521 |
+
feedback_html += f"""
|
| 522 |
+
</div>
|
| 523 |
+
<div style="margin-top: 16px; padding: 12px; background: #dbeafe; border-radius: 8px; text-align: center;">
|
| 524 |
+
<strong style="color: #1e40af;">🎯 Next Goal: Reach {min(accuracy_percentage + 15, 100)}% accuracy!</strong>
|
| 525 |
+
</div>
|
| 526 |
+
</div>
|
| 527 |
+
"""
|
| 528 |
+
|
| 529 |
+
return feedback_html, final_text
|
| 530 |
+
|
| 531 |
+
def assess_student_reading_ui(original_passage_state, student_audio_path):
|
| 532 |
if not student_audio_path: return "🎤 Please record your reading first!", ""
|
| 533 |
if not original_passage_state: return "Hmm, the original story is missing. 😟 Please generate a story first.", ""
|
| 534 |
+
transcribed_text = speech_to_text_whisper_space(student_audio_path)
|
| 535 |
stt_errors = ["couldn't understand", "had trouble", "service isn't working", "service is busy", "didn't get any recording", "filepath type issue"]
|
| 536 |
if any(err in (transcribed_text or "").lower() for err in stt_errors): return transcribed_text, ""
|
| 537 |
+
feedback, highlighted_passage = compare_texts_for_feedback(original_passage_state, transcribed_text)
|
| 538 |
+
return feedback, highlighted_passage
|
| 539 |
|
| 540 |
css = """
|
| 541 |
body, .gradio-container {
|
|
|
|
| 543 |
font-family: -apple-system, BlinkMacSystemFont, 'San Francisco', 'Segoe UI', 'Roboto', Arial, sans-serif !important;
|
| 544 |
}
|
| 545 |
.main-header {
|
| 546 |
+
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%) !important;
|
| 547 |
+
border-radius: 0 !important;
|
| 548 |
+
box-shadow: 0 8px 32px 0 rgba(102, 126, 234, 0.3) !important;
|
| 549 |
+
padding: 32px 20px 28px 20px !important;
|
| 550 |
+
margin: -20px -20px 28px -20px !important;
|
| 551 |
+
width: calc(100% + 40px) !important;
|
| 552 |
text-align: center;
|
| 553 |
border: none !important;
|
| 554 |
+
position: relative;
|
| 555 |
+
overflow: hidden;
|
| 556 |
+
}
|
| 557 |
+
.main-header::before {
|
| 558 |
+
content: '';
|
| 559 |
+
position: absolute;
|
| 560 |
+
top: 0;
|
| 561 |
+
left: 0;
|
| 562 |
+
right: 0;
|
| 563 |
+
bottom: 0;
|
| 564 |
+
background: linear-gradient(45deg, rgba(255,255,255,0.1) 0%, rgba(255,255,255,0.05) 100%);
|
| 565 |
+
pointer-events: none;
|
| 566 |
+
}
|
| 567 |
+
.main-header h1 {
|
| 568 |
+
font-size: 2.4rem !important;
|
| 569 |
+
font-weight: 800 !important;
|
| 570 |
+
color: white !important;
|
| 571 |
+
margin: 0 0 8px 0 !important;
|
| 572 |
+
text-shadow: 0 2px 4px rgba(0,0,0,0.1) !important;
|
| 573 |
+
letter-spacing: -0.5px !important;
|
| 574 |
+
position: relative;
|
| 575 |
+
z-index: 1;
|
| 576 |
+
}
|
| 577 |
+
.main-header p {
|
| 578 |
+
color: rgba(255,255,255,0.9) !important;
|
| 579 |
+
font-size: 1.1rem !important;
|
| 580 |
+
margin: 0 !important;
|
| 581 |
+
font-weight: 400 !important;
|
| 582 |
+
position: relative;
|
| 583 |
+
z-index: 1;
|
| 584 |
+
}
|
| 585 |
+
.tech-badge {
|
| 586 |
+
background: rgba(255,255,255,0.2) !important;
|
| 587 |
+
color: white !important;
|
| 588 |
+
border-radius: 12px !important;
|
| 589 |
+
padding: 4px 12px !important;
|
| 590 |
+
font-size: 12px !important;
|
| 591 |
+
font-weight: 600 !important;
|
| 592 |
+
backdrop-filter: blur(10px) !important;
|
| 593 |
}
|
|
|
|
|
|
|
|
|
|
| 594 |
.gr-block, .gr-panel {background: white !important; border-radius: 18px !important; box-shadow: 0 2px 8px 0 rgba(60,60,90,0.07) !important; border: none !important; padding: 28px 22px !important;}
|
| 595 |
.section-header {background: transparent !important; border: none !important; padding: 0 !important; margin-bottom: 16px !important;}
|
| 596 |
.section-header h3 {color: #1e293b !important; font-size: 1.14rem !important; font-weight: 600 !important;}
|
|
|
|
| 627 |
background: linear-gradient(90deg, #e0e7ef, #dde5f2) !important;
|
| 628 |
color: #2a3140 !important;
|
| 629 |
transition: all 0.15s cubic-bezier(0.4,0.0,0.2,1) !important;
|
| 630 |
+
transform: translateY(0) !important
|
| 631 |
}
|
| 632 |
|
| 633 |
.gr-button[variant="secondary"]:hover {
|
|
|
|
| 703 |
audio_out = gr.Audio(label="🎵 Story Audio", type="filepath", visible=True, autoplay=False)
|
| 704 |
gr.Markdown("""
|
| 705 |
<div style="margin: 20px 0 0 0; padding: 10px 20px; background: #f4f7fa; border-radius: 16px;">
|
| 706 |
+
<b>➡️ Next:</b> Record yourself reading below, then check the "Analysis & Feedback" tab for results.
|
| 707 |
</div>
|
| 708 |
""")
|
| 709 |
stud_audio_in = gr.Audio(sources=["microphone"], type="filepath", label="🎤 Your Recording")
|
|
|
|
| 712 |
assess_btn = gr.Button("🔍 Analyze Reading", variant="primary", size="lg", interactive=False)
|
| 713 |
recording_status = gr.Markdown("", elem_id="recording_status")
|
| 714 |
analysis_status = gr.Markdown("", elem_id="analysis_status")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 715 |
|
| 716 |
+
with gr.TabItem("📊 Analysis & Feedback", elem_id="analysis_tab"):
|
| 717 |
+
with gr.Row():
|
| 718 |
+
with gr.Column(scale=1, variant="panel"):
|
| 719 |
+
gr.Markdown("""
|
| 720 |
+
<div class="section-header">
|
| 721 |
+
<h3>🔍 Word-by-Word Analysis</h3>
|
| 722 |
+
<p>See exactly which words you read correctly</p>
|
| 723 |
+
</div>
|
| 724 |
+
""")
|
| 725 |
+
highlighted_out = gr.HTML(
|
| 726 |
+
value="""
|
| 727 |
+
<div style="text-align: center; color: #6b7280; padding: 20px;">
|
| 728 |
+
<h4>🎯 Detailed Word Analysis</h4>
|
| 729 |
+
<p>Color-coded word analysis will appear here.</p>
|
| 730 |
+
<div style="margin-top: 15px; padding: 15px; background: #f8fafc; border-radius: 12px;">
|
| 731 |
+
<div style="display: flex; flex-wrap: wrap; gap: 8px; justify-content: center; font-size: 0.8rem; margin-bottom: 10px;">
|
| 732 |
+
<span style="background: #22c55e; color: white; padding: 2px 8px; border-radius: 4px;">✓ Correct</span>
|
| 733 |
+
<span style="background: #ef4444; color: white; padding: 2px 8px; border-radius: 4px;">✗ Missed</span>
|
| 734 |
+
<span style="background: #f59e0b; color: white; padding: 2px 8px; border-radius: 4px;">~ Changed</span>
|
| 735 |
+
<span style="background: #8b5cf6; color: white; padding: 2px 8px; border-radius: 4px;">+ Added</span>
|
| 736 |
+
</div>
|
| 737 |
+
<p style="margin: 0; font-size: 14px;">🎤 Complete a reading practice session to see your word analysis!</p>
|
| 738 |
+
</div>
|
| 739 |
+
</div>
|
| 740 |
+
""",
|
| 741 |
+
elem_id="highlighted_passage_output"
|
| 742 |
+
)
|
| 743 |
+
|
| 744 |
+
with gr.Row():
|
| 745 |
+
with gr.Column(scale=1, variant="panel"):
|
| 746 |
+
gr.Markdown("""
|
| 747 |
+
<div class="section-header">
|
| 748 |
+
<h3>📊 Reading Performance</h3>
|
| 749 |
+
<p>Your detailed feedback and scores</p>
|
| 750 |
+
</div>
|
| 751 |
+
""")
|
| 752 |
+
feedback_out = gr.HTML(
|
| 753 |
+
value="""
|
| 754 |
+
<div style="text-align: center; color: #6b7280; padding: 20px;">
|
| 755 |
+
<h4>📈 Performance Analysis</h4>
|
| 756 |
+
<p>Your detailed feedback will appear here after recording.</p>
|
| 757 |
+
<div style="margin-top: 15px; padding: 15px; background: #f8fafc; border-radius: 12px;">
|
| 758 |
+
<p style="margin: 0; font-size: 14px;">💡 <strong>Tip:</strong> Go to the "Practice & Generate" tab to record yourself reading!</p>
|
| 759 |
+
</div>
|
| 760 |
+
</div>
|
| 761 |
+
""",
|
| 762 |
+
elem_id="feedback_output"
|
| 763 |
+
)
|
| 764 |
+
|
| 765 |
with gr.TabItem("ℹ️ About & How It Works", elem_id="about_tab"):
|
| 766 |
gr.Markdown("""
|
| 767 |
<div class="section-header">
|
|
|
|
| 770 |
</div>
|
| 771 |
""")
|
| 772 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 773 |
gr.HTML("""
|
| 774 |
+
<div style="text-align: center; margin: 20px 0;">
|
| 775 |
+
<h3 style="color: #1e293b; margin-bottom: 20px;">📊 Application Workflow</h3>
|
| 776 |
+
<svg width="1400" height="700" xmlns="http://www.w3.org/2000/svg" style="max-width: 100%; height: auto; border: 2px solid #e5e7eb; border-radius: 12px; background: white;">
|
| 777 |
<!-- Background -->
|
| 778 |
<rect width="1400" height="600" fill="#fafafa"/>
|
| 779 |
|
|
|
|
| 869 |
""")
|
| 870 |
|
| 871 |
gr.Markdown("""
|
| 872 |
+
## 🚀 How to Use the App
|
| 873 |
+
|
| 874 |
+
• **Enter details** → Name, grade, and story topic
|
| 875 |
+
• **Generate story** → Click "✨ Generate Story" button
|
| 876 |
+
• **Listen to AI** → Play the audio to hear correct pronunciation
|
| 877 |
+
• **Record yourself** → Use microphone to read the story aloud
|
| 878 |
+
• **Get feedback** → Click "🔍 Analyze Reading" for results
|
| 879 |
+
• **Practice more** → Try new topics and improve your reading!
|
| 880 |
+
|
| 881 |
---
|
| 882 |
|
| 883 |
## 🔧 Key Components
|
|
|
|
| 907 |
passage_state = story_text
|
| 908 |
return story_text, audio_btn_update, audio_player_update, passage_state
|
| 909 |
|
| 910 |
+
def assess_reading_with_analysis(original_passage_state, student_audio_path):
|
| 911 |
if not student_audio_path:
|
| 912 |
return (
|
| 913 |
"""
|
|
|
|
| 929 |
""
|
| 930 |
)
|
| 931 |
|
|
|
|
|
|
|
|
|
|
| 932 |
# Start transcription
|
| 933 |
+
transcribed_text = speech_to_text_whisper_space(student_audio_path)
|
|
|
|
| 934 |
|
| 935 |
stt_errors = ["couldn't understand", "had trouble", "service isn't working", "service is busy", "didn't get any recording", "filepath type issue"]
|
| 936 |
if any(err in (transcribed_text or "").lower() for err in stt_errors):
|
|
|
|
| 945 |
""
|
| 946 |
)
|
| 947 |
|
|
|
|
| 948 |
feedback, highlighted_passage = compare_texts_for_feedback(original_passage_state, transcribed_text)
|
|
|
|
| 949 |
|
| 950 |
analysis_msg = """
|
| 951 |
<div class="status-indicator status-success">
|
|
|
|
| 955 |
"""
|
| 956 |
return (analysis_msg, feedback, highlighted_passage)
|
| 957 |
|
| 958 |
+
def generate_story_and_audio_complete(name, grade, topic):
|
| 959 |
+
"""Generate story and audio in one function - Gradio will show default loading indicators"""
|
| 960 |
+
|
| 961 |
+
# Generate story text first
|
| 962 |
+
story_text = generate_story_from_llm(name, grade, topic)
|
| 963 |
+
if not story_text:
|
| 964 |
+
return "", None, ""
|
| 965 |
+
|
| 966 |
+
# Check if story generation was successful
|
| 967 |
+
if not story_text or any(err in story_text.lower() for err in ["error", "blocked", "couldn't", "api key not configured"]):
|
| 968 |
+
return story_text, None, story_text
|
| 969 |
+
|
| 970 |
+
# Generate audio (Gradio will show its loading indicator)
|
| 971 |
+
audio_filepath = text_to_speech_using_space_simple(story_text)
|
| 972 |
+
|
| 973 |
+
if audio_filepath:
|
| 974 |
+
print(f"COMPLETE: Story and audio generated successfully")
|
| 975 |
+
return story_text, audio_filepath, story_text
|
| 976 |
+
else:
|
| 977 |
+
print("COMPLETE: Story generated, but audio failed")
|
| 978 |
+
return story_text, None, story_text
|
| 979 |
+
|
| 980 |
def update_recording_status(audio_file):
|
| 981 |
if audio_file is not None:
|
| 982 |
return (
|
|
|
|
| 1048 |
gr.update(visible=False),
|
| 1049 |
gr.update(interactive=False)
|
| 1050 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1051 |
|
| 1052 |
+
# Event handlers - Generate story first, then audio separately
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1053 |
gen_btn.click(
|
| 1054 |
+
fn=generate_story_from_llm,
|
| 1055 |
inputs=[s_name, s_grade, s_topic],
|
| 1056 |
+
outputs=[passage_out]
|
| 1057 |
+
).then(
|
| 1058 |
+
fn=lambda story_text: story_text, # Store story in state immediately
|
| 1059 |
+
inputs=[passage_out],
|
| 1060 |
+
outputs=[original_passage_state]
|
| 1061 |
+
)
|
| 1062 |
+
|
| 1063 |
+
# Separate audio generation triggered by story output change
|
| 1064 |
+
passage_out.change(
|
| 1065 |
+
fn=lambda story_text: text_to_speech_using_space_simple(story_text) if story_text and not any(err in story_text.lower() for err in ["error", "blocked", "couldn't", "api key not configured"]) else None,
|
| 1066 |
+
inputs=[passage_out],
|
| 1067 |
+
outputs=[audio_out]
|
| 1068 |
)
|
| 1069 |
|
| 1070 |
assess_btn.click(
|
| 1071 |
fn=assess_reading_with_analysis,
|
| 1072 |
inputs=[original_passage_state, stud_audio_in],
|
| 1073 |
+
outputs=[analysis_status, feedback_out, highlighted_out]
|
|
|
|
| 1074 |
)
|
| 1075 |
|
| 1076 |
stud_audio_in.change(
|