obstinate 发表于 2025-3-28 18:03:28
http://reply.papertrans.cn/25/2424/242337/242337_41.png不公开 发表于 2025-3-28 19:38:22
http://reply.papertrans.cn/25/2424/242337/242337_42.png古董 发表于 2025-3-28 23:36:01
http://reply.papertrans.cn/25/2424/242337/242337_43.pngdeforestation 发表于 2025-3-29 06:43:18
,Characterizing Model Robustness via Natural Input Gradients,without complex adversarial optimization. Our analyses also highlight the relationship between model robustness and properties of natural input gradients, such as asymmetric sample and channel statistics. Surprisingly, we find model robustness can be significantly improved by simply regularizing itsN斯巴达人 发表于 2025-3-29 09:12:27
http://reply.papertrans.cn/25/2424/242337/242337_45.pngTEN 发表于 2025-3-29 12:07:14
http://reply.papertrans.cn/25/2424/242337/242337_46.png分期付款 发表于 2025-3-29 17:29:12
,Tuning-Free Image Customization with Image and Text Guidance, utilizes text and image guidance for image customization in specific regions. Our approach outperforms previous methods in both human and quantitative evaluations, providing an efficient solution for various practical applications, such as image synthesis, design, and creative photography. Project声明 发表于 2025-3-29 19:57:27
,FairDomain: Achieving Fairness in Cross-Domain Medical Image Segmentation and Classification,prove fairness by using self-attention to adjust feature importance based on demographic attributes. Additionally, we curate the first fairness-focused dataset with two paired imaging modalities for the same patient cohort on medical segmentation and classification tasks, to rigorously assess fairne不透气 发表于 2025-3-30 02:50:02
,Emerging Property of Masked Token for Effective Pre-training,l approach termed ., specifically designed to improve model efficiency through weight recalibration and the enhancement of the key property of masked tokens. The proposed method serves as an adaptable solution that seamlessly integrates into any MIM approach that leverages masked tokens. As a resultmastopexy 发表于 2025-3-30 04:59:20
,Track2Act: Predicting Point Tracks from Internet Videos Enables Generalizable Robot Manipulation,g residual actions through a closed loop policy trained with a few embodiment-specific demonstrations. We show that this approach of combining scalably learned track prediction with a residual policy requiring minimal in-domain robot-specific data enables diverse generalizable robot manipulation, an