[论文速览]EViT: Expediting Vision Transformers via Token Reorganizations[2202.07800]
作者: 秋刀鱼的炼丹工坊
作者简介: 经中此篇如此高深，我确实不懂。
描述: 论文题目:Not All Patches are What You Need: Expediting Vision Transformers via Token Reorganizations 论文地址:http://arxiv.org/abs/2202.07800 代码: https://github.com/youweiliang/evit TokenLearner: BV1of4y1t77m * 本视频旨在传递一篇论文的存在推荐感兴趣的您阅读，并不是详细介绍，受up能力限制经常出现中英混杂，散装英语等现象，请见谅。如论文报道出了偏差，欢迎各位怒斥。 ** 新论文推荐，过往论文查找，欢迎编辑这个文档： https://docs.qq.com/sheet/DSUdOTG9xWUdydVB6 *** Slides每1-2月会上传到置顶动态地址
![[论文速览]EViT: Expediting Vision Transformers via Token Reorganizations[2202.07800]](/image/cab99e2b1d0ddc5415d6582fd3686ff3a1b0e07d.jpg)
![[论文夕拾]Diffusion Models for Robotics](/image/c8225b2e0a5ad69287831ab08d1663f7f213d748.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]OpenVLA: An Open-Source Vision-Language-Action Model[2406.09246]](/image/97b2ad4dadd7714ad58325530a1aec7346b60b63.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Visual Prompt Tuning / VPT[2203.12119]](/image/2e914ba575599f48de14cf4a0f954bd5f01d98a2.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]LLaMA-Adapter: Efficient Fine-tuning..Zero-init Attention[2303.16199]](/image/ff64905ff76efb4bdc968dba0fb20828d5ee895f.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]MViT: Multiscale Vision Transformers[2104.11227]](/image/76601dbb2bc4001e5c21ad930219dd7750550b12.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Theia: Distilling Diverse Vision Foundation Models for Robot..[2407.20179]](/image/688292fe9375bb0fa280c7ecaf87c5aba6d79d4a.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]LLaVA: Visual Instruction Tuning[2304.08485]](/image/e70f51855bd470691b72d85181c90bd4742e9107.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Masked-attention Mask Tr. for Universal Image Segmentation[2112.01527]](/image/9340a95ed9b82a90bc26ba6f5d4ba8be4964086e.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Efficient Action Tokenization for Vision-Language-Action Model[2501.09747]](/image/282197e006141735727e8b3523e3b77d5075f121.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]RegMixup: Mixup as a Regularizer Can Surprisingly Improve...[2206.14502]](/image/77d1c97a4cd944e19ad822be054d6a72dd9d1ca6.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]LoRA: Low-Rank Adaptation of Large Language Models[2106.09685]](/image/2f4cb362cc5e24a347a012cfa05220728950cc9a.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]Location-Aware Self-Supervised Transformers for Semantic Seg.[2212.02400]](/image/6b3596bfeeefb31a9da5e8c44de13897ccfaa33e.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]DeiT: Data-efficient Image Transformers[2012.12877]](/image/dcc355c0e181bc0795eb94f32bc610c619a3644f.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Ferret-v2: An Improved...for Referring and Grounding with LLMs[2404.07973]](/image/27c3c75e59cd3515af31bc1c99d267c2092bd4cb.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Learning to Learn with Generative Models of NN Checkpoints[2209.12892]](/image/ae47e9b9f347382b250a0bfa71800e38227f4104.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]DDPG&TD3[1509.02971][1802.09477]](/image/1e1512d8675a6dff55aadac90b3a47a15c16e361.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]Visual Autoregressive Modeling: ...via Next-Scale Prediction[2404.02905]](/image/31db23ff9328d0681390bf96ac59078fcd000824.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]Transf. Meta-learners for Implicit Neural Representations[2208.02801]](/image/e5079a60b6d6d21be0f4406097be868ddc7baf76.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Bootstrapping Language-Image Pre-training...[2201.12086]](/image/49d17247debc220e9ed7092ebc91e1197e08359b.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]DINO Emerging Properties in SelfSupervised Vision Transformers[2104.14294]](/image/ec6eaec1f853d6aad9a3a04299962364a2069c9a.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Ferret: Refer and Ground Anything Anywhere at Any Granularity[2310.07704]](/image/01f1c66b1653d823d3f9e575abb202d877134de0.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]DAT: Vision Transformer with Deformable Attention[2201.00520]](/image/e82720fd9f531018e32df9ea33690cb799e92c57.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]NeRF-RL: Reinforcement Learning with Neural Radiance Fields[2206.01634]](/image/e7657581954d0419fe11141cc9f2cb139c03ae3a.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]Deconstructing Denoising Diffusion Models for SSL[2401.14404]](/image/a1fa1e1345d3cc4b48a751183547e4b36ee83bde.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]SODA: Bottleneck Diffusion Models for Representation Learning[2311.17901]](/image/b973a0cc9e9c340ffe3734a81f8287ca5de2dbfc.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]ViperGPT: Visual Inference via Python Execution for Reasoning[2303.08128]](/image/06a1c2f6646dbfe34f4c0bdec2d3f24b40227b69.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Aggregating Nested Transformers[2105.12723]](/image/d421d14158808a20d6b8f85141adce23fcf59e4a.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Self-supervised Video Transformer[2112.01514]](/image/bf6057e7f81b327d4709b50c7d02a2f3b72c2001.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文简析]Energy-Inspired Self-Supervised Pretraining for Vision Models[2302.01384]](/image/9055d4505c9bb83eddacda6ed9b87839ca0977b2.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]iBOT: Image BERT Pre-Training with Online Tokenizer[2111.07832]](/image/e09695eb16e634c5d85221d07c3ab19e5d1ffb6e.jpg@336w_190h_!web-video-rcmd-cover.avif)
![[论文速览]Structured Denoising Diffusion Models in Discrete State-Spaces[2107.03006]](/image/e49c035455c73db15a4b53ba4d8a20c02fbdc210.jpg@336w_190h_!web-video-rcmd-cover.avif)