search

megvii-model / YOLOF

MIT License
809 stars 115 forks source link

paper #4

Closed abcxs closed 3 years ago

abcxs commented 3 years ago

hello, the effect of large objects is a 9.3% difference, which is not small compared to others image image

abcxs commented 3 years ago

Sorry, I see what you mean. However, if you compare the first line to the third, it's still a big gap

chensnathan commented 3 years ago

Thanks for your interest.

Yes. Uniform Matching has a big effect on objects across various scales when we apply it to the baseline (the first line). It makes the model be fully trained as all ground truths participate in the training phase. However, its effect on large objects diminishes when we add Dilated Encoder. Thus, we may think that Uniform Matching improves the performance of small and medium objects largely while Dilated Encoder dominates large objects' performance.

abcxs commented 3 years ago

thanks