[논문 정리] Object-Part Attention Model for Fine-grained Image Classification

논문정보 Object-Part Attention Model for Fine-grained Image Classification Object-Part Attention Model for Fine-grained Image Classification Fine-grained image classification is to recognize hundreds of subcategories belonging to the same basic-level category, such as 200 subcategories belonging to the bird, which is highly challenging due to large variance in the same subcategory and small var arxi..

• format_list_bulleted 논문
• · 2023. 4. 18.
• textsms

[python] heatmap - gradcam

모델의 activation visualization 부분을 분석해보자. simple_grad_cam def simple_grad_cam(features, classifier, target_class): """ calculate gradient map. """ features = nn.Parameter(features) logits = torch.matmul(features, classifier) logits[0, :, :, target_class].sum().backward() features_grad = features.grad[0].sum(0).sum(0).unsqueeze(0).unsqueeze(0) gramcam = F.relu(features_grad * features[0]) gramcam =..

• format_list_bulleted 프로그래밍/파이썬 Python
• · 2023. 4. 11.
• textsms

2023-03-27~2023-03-31

보호되어 있는 글입니다.

• format_list_bulleted 연구노트/개발사항
• · 2023. 3. 28.

WeaklySelector

보호되어 있는 글입니다.

• format_list_bulleted 연구노트/관련 자료 정리
• · 2023. 3. 23.

FPN

보호되어 있는 글입니다.

• format_list_bulleted 연구노트/관련 자료 정리
• · 2023. 3. 22.

GCN combiner

보호되어 있는 글입니다.

• format_list_bulleted 연구노트/관련 자료 정리
• · 2023. 3. 22.