A Novel Transformer based Semantic Segmentation Scheme for Fine-Resolution Remote Sensing Images（語義分割任務(wù)）

Self-Supervised Learning with Swin Transformers（模型簡稱：MoBY，使用了對比學(xué)習(xí)）

Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation（醫(yī)療圖像語義分割）

Rethinking Training from Scratch for Object Detection（看不懂）

Demystifying Local Vision Transformer: Sparse Connectivity, Weight Sharing, and Dynamic Weight


DS-TransUNet: Dual Swin Transformer U-Net for Medical Image Segmentation（醫(yī)療圖像的語義分割）

Long-Short Temporal Contrastive Learning of Video Transformers

Video Swin Transformer



PVTv2: Improved Baselines with Pyramid Vision Transformer（Pyramid：金字塔）


CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows



What Makes for Hierarchical Vision Transformer?

CYCLEMLP: A MLP-LIKE ARCHITECTURE FOR DENSE PREDICTION


Congested Crowd Instance Localization with Dilated Convolutional Swin Transformer

ConvNets vs. Transformers: Whose Visual Representations are More Transferable?

Vision transformers have attracted much attention from computer vision researchers as they are not restricted to the spatial inductive bias of ConvNets. However, although Transformer-based backbones have achieved much progress on ImageNet classification, it is still unclear whether the learned representations are as transferable as or even more transferable than ConvNets’ features. To address this point, we systematically investigate the transfer learning ability of ConvNets and vision transformers in 15 single-task and multi-task performance evaluations. Given the strong correlation between the performance of pretrained models and transfer learning, we include 2 residual ConvNets (i.e., R-101×3 and R-152×4) and 3 Transformer based visual backbones (i.e., ViT-B, ViT-L and Swin-B), which have close error rates on ImageNet, that indicate similar transfer learning performance on downstream datasets. We observe consistent advantages of Transformer-based backbones on 13 downstream tasks (out of 15), including but not limited to fine-grained classification, scene recognition (classification, segmentation and depth estimation), open-domain classification, face recognition, etc. More specifically, we find that two ViT models heavily rely on whole network fine-tuning to achieve performance gains while Swin Transformer does not have such a requirement. Moreover, vision transformers behave more robustly in multi-task learning, i.e., bringing more improvements when managing mutually beneficial tasks and reducing performance losses when tackling irrelevant tasks. We hope our discoveries can facilitate the exploration and exploitation of vision transformers in the future.
視覺變壓器因其不局限于卷積神經(jīng)網(wǎng)絡(luò)的空間感應(yīng)偏置而受到計算機視覺研究者的廣泛關(guān)注。然而，盡管基于transformer的主干網(wǎng)在ImageNet分類方面取得了很大的進展，但我們?nèi)匀徊磺宄W(xué)習(xí)后的表示是否和卷積網(wǎng)絡(luò)的特征一樣可轉(zhuǎn)移，甚至比卷積網(wǎng)絡(luò)的特征更可轉(zhuǎn)移。為了解決這一問題，我們在15個單任務(wù)和多任務(wù)性能評估中系統(tǒng)地研究了卷積神經(jīng)網(wǎng)絡(luò)和視覺變壓器的遷移學(xué)習(xí)能力。考慮到預(yù)訓(xùn)練模型的性能與遷移學(xué)習(xí)之間的強相關(guān)性，我們包括2個殘差ConvNets (R-101×3和R-152×4)和3個基于Transformer的視覺主干(vi - b、vi - l和swi - b)，它們在ImageNet上的錯誤率接近，表明在下游數(shù)據(jù)集上的遷移學(xué)習(xí)性能類似。我們觀察到基于transformer的骨干在13個下游任務(wù)(15個任務(wù)中)上具有一致的優(yōu)勢，包括但不限于細(xì)粒度分類、場景識別(分類、分割和深度估計)、開放域分類、人臉識別等。更具體地說，我們發(fā)現(xiàn)兩個ViT模型嚴(yán)重依賴于整個網(wǎng)絡(luò)的微調(diào)來實現(xiàn)性能增益，而Swin Transformer沒有這樣的需求。此外，視覺變壓器在多任務(wù)學(xué)習(xí)中表現(xiàn)得更加穩(wěn)健，即在管理互惠任務(wù)時帶來更多的改進，在處理無關(guān)任務(wù)時減少性能損失。我們希望我們的發(fā)現(xiàn)可以促進未來視覺變壓器的探索和開發(fā)。

SwinIR: Image Restoration Using Swin Transformer（重點是殘差鏈接）

Sparse MLP for Image Recognition: Is Self-Attention Really Necessary?

Semi-Supervised Wide-Angle Portraits Correction by Multi-Scale Transformer



3rd Place Scheme on Instance Segmentation Track of ICCV 2021 VIPriors Challenges

VIDT: AN EFFICIENT AND EFFECTIVE FULLY TRANSFORMER-BASED OBJECT DETECTOR


Satellite Image Semantic Segmentation（衛(wèi)星圖像語義分割）（手稿）
COVID-19 Detection in Chest X-ray Images Using Swin Transformer and Transformer in Transformer

HRFormer: High-Resolution Transformer for Dense Prediction


Vis-TOP: Visual Transformer Overlay Processor





Hepatic vessel segmentation based on 3D swin-transformer with inductive biased multi-head self-attention


Transformer-based Image Compression（圖像壓縮）



Swin Transformer V2: Scaling Up Capacity and Resolution

Vision Transformer with Deformable Attention

Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images

SWIN-POSE: SWIN TRANSFORMER BASED HUMAN POSE ESTIMATION

總結(jié)

以上是生活随笔為你收集整理的关于大家对Swin Transformer的魔改论文模型记录（只关注Swin是如何使用的）的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯，歡迎將生活随笔推薦給好友。