전체 글88 Pytorch Learnable parameter, 학습가능한 파라미터 만들기 우리가 pytorch 내에서 학습 가능한 파라미터를 만들어야 할 때가 종종 있다. 1 이때 nn.Module 내에 파라미터를 선언할 경우 parameter = nn.Parameter(torch.randn(10).cuda()) 위와 같이 선언하여 사용하면 된다. 2 nn.Module 외에서 위 파라미터를 선언한 후 아래와 같이 연산을 한다면 parameter = parameter.exp() parameter = parameter.double() 아래와 같은 ValueError: can't optimize a non-leaf Tensor 에러를 만나게 된다. 이는 이미 만들어진 파라미터를 다른연산으로 casting하였기 때문이라고 하는데 pytorch를 자세히 공부하지 않아서 이해는 안된다. 하지만 해결 방.. 2024. 3. 14. [논문리뷰] Transductive Zero-Shot Learning for 3D Point Cloud Classification Paper OverviewWACV'20https://arxiv.org/abs/1912.07161 Transductive Zero-Shot Learning for 3D Point Cloud ClassificationZero-shot learning, the task of learning to recognize new classes not seen during training, has received considerable attention in the case of 2D image classification. However despite the increasing ubiquity of 3D sensors, the corresponding 3D point cloudarxiv.org Abstract이 논문은 .. 2024. 3. 12. [논문리뷰] Generalized Zero-Shot Learning Via Over-Complete Distribution Paper Overview CVPR'20 https://arxiv.org/abs/2004.00666 Generalized Zero-Shot Learning Via Over-Complete Distribution A well trained and generalized deep neural network (DNN) should be robust to both seen and unseen classes. However, the performance of most of the existing supervised DNN algorithms degrade for classes which are unseen in the training set. To learn a discr arxiv.org Abstract Zero.. 2024. 3. 11. [논문리뷰] Mitigating the Hubness Problem for Zero-Shot Learning of 3D Objects Paper OverviewBMVC'19https://arxiv.org/abs/1907.06371 Mitigating the Hubness Problem for Zero-Shot Learning of 3D ObjectsThe development of advanced 3D sensors has enabled many objects to be captured in the wild at a large scale, and a 3D object recognition system may therefore encounter many objects for which the system has received no training. Zero-Shot Learning (ZSL) apparxiv.org Abstract3D .. 2024. 3. 5. [논문리뷰] HSVA: Hierarchical Semantic-Visual Adaptation for Zero-Shot Learning Paper Overview NeurIPS'21 https://arxiv.org/abs/2109.15163 HSVA: Hierarchical Semantic-Visual Adaptation for Zero-Shot Learning Zero-shot learning (ZSL) tackles the unseen class recognition problem, transferring semantic knowledge from seen classes to unseen ones. Typically, to guarantee desirable knowledge transfer, a common (latent) space is adopted for associating the visual and arxiv.org Abs.. 2024. 2. 27. [논문리뷰] Open-world Semi-supervised Novel Class Discovery Paper Overview IJCAI'23 https://arxiv.org/abs/2305.13095 Open-world Semi-supervised Novel Class Discovery Traditional semi-supervised learning tasks assume that both labeled and unlabeled data follow the same class distribution, but the realistic open-world scenarios are of more complexity with unknown novel classes mixed in the unlabeled set. Therefore, it is arxiv.org Abstract 전통적인 semi-superv.. 2024. 2. 14. 이전 1 2 3 4 5 6 7 8 ··· 15 다음
