Transform the subject from the input image into appealing claymation character, keep likeness. Handmade plasticine style, ...
Abstract: Many images and videos are primarily processed by computer vision algorithms, involving only occasional human inspection. When this content requires compression before processing, e.g., in ...
Abstract: Currently, deep learning-based joint source channel coding (JSCC) methods have achieved significant progress in enabling semantic communication. However, existing methods of this type often ...
Official repository for the AAAI2025 paper Can We Get Rid of Handcrafted Feature Extractors? SparseViT: Nonsemantics-Centered, Parameter-Efficient Image Manipulation Localization through Spare-Coding ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results