Notes from the work
Knowledge Base
What I learn while building with ML and agents, written up so it is useful to the next person who hits the same problem: where the standard approach breaks, how others fix it, and what I measured myself, or a plain "not measured yet".
- Does INT8 quantization hurt small object detection? The damage mAP does not show
Quantizing a detector to INT8 costs small objects 1.4 to 3 times more than the overall metric, and below 12 pixels the published loss reaches a third of the accuracy; what is known, the fixes, and what nobody has measured yet.
By concept
- post-training-quantization
- small-object-detection