20259dcad0
- Add Honda client with static_general and video_general profiles - Add video QC capability using Gemini native video analysis (4 checks: visual_quality, brand_consistency, text_legibility, pacing_flow) - Add video_general profile assigned to all 8 clients - Extend session lifetime with MSAL silent token refresh (proactive every 45min + reactive on expiry), switch cache to localStorage - Re-enable OCR layout measurements for Amazon checks - Add scope boundary notes to all 6 Amazon checks to prevent cross- check penalization (locale errors isolated to logo_country only) - Relax margins left-alignment tolerance from 1% to 4% to account for logo lockup internal padding - Update brand guidelines DB with Amazon localization matrix and processed Dove PDF summary Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
103 lines
4.3 KiB
Python
103 lines
4.3 KiB
Python
import os
|
|
import sys
|
|
|
|
# Add parent directory to path to import shared modules
|
|
sys.path.append(os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))))
|
|
|
|
from visual_qc_apps.flask_app_template import FlaskAppTemplate
|
|
|
|
class VideoTextLegibilityApp(FlaskAppTemplate):
|
|
"""
|
|
Video QC - Text & Subtitle Legibility Check
|
|
Evaluates all on-screen text for readability, timing, and accessibility.
|
|
"""
|
|
|
|
def __init__(self):
|
|
prompt = """You are performing a text legibility quality-control check on a VIDEO marketing asset. You are watching the full video — evaluate ALL on-screen text throughout the entire duration.
|
|
|
|
VIDEO TEXT LEGIBILITY EVALUATION:
|
|
|
|
Evaluate the following for every text element that appears in the video:
|
|
|
|
1. TEXT READABILITY:
|
|
- Is all on-screen text large enough to read comfortably?
|
|
- Is there sufficient contrast between text and background?
|
|
- Are text elements given enough screen time to be fully read?
|
|
- Rule of thumb: viewers need approximately 1 second per 3 words to comfortably read text
|
|
|
|
2. TEXT CONTENT ACCURACY:
|
|
- Are there any visible spelling mistakes or typos in on-screen text?
|
|
- Are dates, prices, URLs, or other factual information formatted correctly?
|
|
- Is the language consistent throughout (no unexpected language switches)?
|
|
- Are any words truncated or cut off by the frame edges?
|
|
|
|
3. SUBTITLE / CAPTION ASSESSMENT (if present):
|
|
- Are subtitles/captions timed correctly with the spoken content?
|
|
- Are subtitles readable against the video background?
|
|
- Is the subtitle font size appropriate?
|
|
- Are subtitles positioned so they don't obscure important visual content?
|
|
|
|
4. LEGAL / DISCLAIMER TEXT:
|
|
- If legal text or disclaimers are present, are they readable?
|
|
- Is the legal text displayed for an adequate duration?
|
|
- Note: Legal text is often intentionally small, but must still be technically readable
|
|
|
|
5. CALL-TO-ACTION (CTA) TEXT:
|
|
- If a CTA is present (URL, "Shop Now", "Learn More", etc.), is it clearly visible?
|
|
- Is the CTA displayed long enough for viewers to act on it?
|
|
- Is the CTA positioned prominently?
|
|
|
|
STEPS TO EVALUATE:
|
|
1. Watch the full video and note every text element that appears on screen
|
|
2. For each text element, assess readability (size, contrast, duration)
|
|
3. Check for any spelling errors, typos, or formatting issues
|
|
4. Evaluate subtitle timing and readability if present
|
|
5. Check legal text and CTA visibility
|
|
6. Note any text that is truncated, obscured, or appears too briefly
|
|
|
|
DECISION CRITERIA:
|
|
- PASS (score 7-10): All text is readable, correctly spelled, displayed long enough, and has good contrast
|
|
- MODERATE ISSUES (score 4-6): Minor readability issues — slightly small text, brief display duration, or minor contrast issues
|
|
- FAIL (score 1-3): Significant text issues — unreadable text, spelling errors, text too brief to read, or poor contrast making text invisible
|
|
|
|
YOUR OUTPUT:
|
|
Format your response as JSON:
|
|
{
|
|
"text_legibility_check": "Pass" or "Fail",
|
|
"text_elements_found": ["List of text elements observed in the video"],
|
|
"readability": {
|
|
"overall": "Good" or "Moderate" or "Poor",
|
|
"contrast_issues": ["List any text with poor contrast, or empty array"],
|
|
"too_brief": ["List any text that appears too briefly to read, or empty array"],
|
|
"too_small": ["List any text that appears too small, or empty array"]
|
|
},
|
|
"content_accuracy": {
|
|
"spelling_errors": ["List any spelling/typo issues found, or empty array"],
|
|
"formatting_issues": ["List any formatting problems, or empty array"],
|
|
"language_consistency": true or false
|
|
},
|
|
"subtitles": {
|
|
"present": true or false,
|
|
"timing_sync": "Good" or "Slightly off" or "Poor" or "N/A",
|
|
"readable": true or false or "N/A"
|
|
},
|
|
"legal_text": {
|
|
"present": true or false,
|
|
"readable": true or false or "N/A",
|
|
"adequate_duration": true or false or "N/A"
|
|
},
|
|
"cta": {
|
|
"present": true or false,
|
|
"visible": true or false or "N/A",
|
|
"adequate_duration": true or false or "N/A"
|
|
},
|
|
"explanation": "Detailed reasoning for the overall assessment",
|
|
"recommendations": ["List specific recommendations if applicable, else an empty array"]
|
|
}"""
|
|
|
|
super().__init__(__name__, prompt)
|
|
|
|
# Run the app if executed directly
|
|
if __name__ == "__main__":
|
|
app_instance = VideoTextLegibilityApp()
|
|
app_instance.run()
|