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 |
---|---|---|---|---|
244 | PHP7 으로 업그레이드 | WHRIA | 2016.05.05 | 136 |
243 | fast rcnn | WHRIA | 2017.01.08 | 133 |
242 | high impact journals | WHRIA | 2019.04.07 | 132 |
241 | 이미지 비교 [1] | WHRIA | 2019.04.23 | 131 |
240 | KMA 연수교육 주소 | WHRIA | 2016.10.05 | 129 |
239 | caffe | WHRIA | 2016.10.19 | 127 |
238 | popup | WHRIA | 2020.12.31 | 126 |
237 | MXNET SSD | WHRIA | 2019.09.04 | 122 |
236 | distributed training [2] | WHRIA | 2019.10.30 | 122 |
235 | opencv rotate | WHRIA | 2019.03.09 | 120 |
234 | 윈도우 파일 검색 옵션 | WHRIA | 2019.06.14 | 120 |
233 | scx 4623f reset / memory [2] | WHRIA | 2016.01.23 | 117 |
232 | 이미지 size 줄이는 batch | WHRIA | 2016.10.19 | 116 |
231 | MS bot framework [3] | WHRIA | 2020.06.13 | 116 |
230 | 인정이자 | WHRIA | 2016.01.23 | 115 |