- Simplified language for better readability - Added specific performance expectations (4-5 FPS) - Clear test scenarios for QA validation - Direct problem-to-solution mapping - Removed technical jargon for broader audience 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>
1.2 KiB
1.2 KiB
Release Notes - Cluster4NPU
v0.0.2
Release Date: July 31, 2025
Issues Resolved
- App no longer crashes when running inference multiple times
- FPS numbers now show correctly - no more jumping from high to low values
- System stays fast during long sessions - memory usage is now controlled
- Processing won't get stuck when too many images pile up
Improvements
- FPS Display: Shows real processing speed (around 4-5 FPS typically)
- Memory Usage: App automatically cleans up old data to stay efficient
- Real-time Processing: Always processes the newest camera images first
- Removed unnecessary text and confusing messages that cluttered the display
For QA
Test These Scenarios:
- Multiple Sessions: Start and stop inference 3-5 times → Should not crash
- Long Running: Let it run for 1 hour → Performance should stay consistent
- FPS Check: Watch FPS numbers → Should be steady around 4-5 FPS, not jumping wildly
- Memory Check: Monitor system memory → Should not keep growing over time
Expected Results:
- No application crashes
- Stable FPS readings
- Memory usage stays controlled
- System responds quickly to new camera frames