2016 ilsvrc
2019.08.31 21:20
Trimps-Soushen used the pretrained models from Inception-v3, Inception-v4, Inception-ResNet-v2, Pre-Activation ResNet-200, and Wide ResNet (WRN-68–2) for classification, and found out Top-10 difficult categories as above. Diverse results are obtained, which means there is no models being dominant for all categories. Each of the models are strong at classifying some categories, but also weak at classifying some categories. The diversity of models can be used for improving the accuracy. During training, Trimps-Soushen just performed multi-scale augmentation & large mini batch size. During testing, multi-scale + flip are used with dense fusion.
댓글 0
번호 | 제목 | 글쓴이 | 날짜 | 조회 수 |
---|---|---|---|---|
1699 | MedicalPhoto MSVC 2015 와 최신 boost 로... | WHRIA | 2016.08.26 | 47 |
1698 | melanoma awareness [1] | WHRIA | 2019.12.27 | 47 |
1697 | batch normalize | WHRIA | 2020.02.06 | 47 |
1696 | android / capture and crop [5] | WHRIA | 2020.05.10 | 47 |
1695 | sql 중복제거 | WHRIA | 2020.02.10 | 48 |
1694 | conflict of interest | WHRIA | 2020.05.03 | 48 |
1693 | sql replication 성공 [1] | WHRIA | 2020.02.20 | 49 |
1692 | Visual Studio 설치후 Excel 2002 (XP) 종료시 에러 문제 | WHRIA | 2015.11.20 | 50 |
1691 | multiple NIC ubuntu [6] | WHRIA | 2020.03.10 | 50 |
1690 | CE | WHRIA | 2020.05.17 | 50 |
1689 | DDR-WRT IPTV | WHRIA | 2016.06.15 | 51 |
1688 | 너무도 많이 변해버린 c++ | WHRIA | 2016.08.15 | 51 |
1687 | nvidia docker | WHRIA | 2020.01.15 | 51 |
1686 | 감가상각 계산 | WHRIA | 2015.11.30 | 52 |
1685 | nvidia caffe | WHRIA | 2020.01.12 | 52 |