Awesome Transformer

A curated list of all kinds of transformers, also include some personal experiment results, applications and thoughts from industry.

Updates

• 2021.02.20: I opened github discuss panel, we can start discuss about transformers there.

Blogs

• Attention is All You Need
• [Chinese Blog] 3W字长文带你轻松入门视觉transformer [Link]
• Transformers in Vision: A Survey [paper] - 2021.01.04
• A Survey on Visual Transformer [paper] - 2020.12.24
• https://zhuanlan.zhihu.com/p/342512339 万字长文带你了解2021大热的ViT
• 线性Attention的探索：Attention必须有个Softmax吗？link

Standalone Github Repos

• https://github.com/ThilinaRajapakse/simpletransformers: Transformers for Classification, NER, QA, Language Modelling, Language Generation, T5, Multi-Modal, and Conversational AI;

arXiv papers

• Training Vision Transformers for Image Retrieval[paper]
• [TransReID] TransReID: Transformer-based Object Re-Identification[paper]
• [VTN] Video Transformer Network[paper]
• [T2T-ViT] Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet [paper] [code]
• [BoTNet] Bottleneck Transformers for Visual Recognition [paper]
• [CPTR] CPTR: Full Transformer Network for Image Captioning [paper]
• Learn to Dance with AIST++: Music Conditioned 3D Dance Generation [paper] [code]
• [Trans2Seg] Segmenting Transparent Object in the Wild with Transformer [paper] [code]
• [SMCA] Fast Convergence of DETR with Spatially Modulated Co-Attention [paper]
• Investigating the Vision Transformer Model for Image Retrieval Tasks [paper]
• [Trear] Trear: Transformer-based RGB-D Egocentric Action Recognition [paper]
• [VisTR] End-to-End Video Instance Segmentation with Transformers [paper]
• [VisualSparta] VisualSparta: Sparse Transformer Fragment-level Matching for Large-scale Text-to-Image Search [paper]
• [TrackFormer] TrackFormer: Multi-Object Tracking with Transformers [paper]
• [LETR] Line Segment Detection Using Transformers without Edges [paper]
• [TAPE] Transformer Guided Geometry Model for Flow-Based Unsupervised Visual Odometry [paper]
• [TRIQ] Transformer for Image Quality Assessment [paper] [code]
• [TransTrack] TransTrack: Multiple-Object Tracking with Transformer [paper] [code]
• [SETR] Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers [paper] [code]
• [TransPose] TransPose: Towards Explainable Human Pose Estimation by Transformer [paper]
• [DeiT] Training data-efficient image transformers & distillation through attention [paper]
• [Pointformer] 3D Object Detection with Pointformer [paper]
• [ViT-FRCNN] Toward Transformer-Based Object Detection [paper]
• [Taming-transformers] Taming Transformers for High-Resolution Image Synthesis [paper] [code]
• [SceneFormer] SceneFormer: Indoor Scene Generation with Transformers [paper]
• [PCT] PCT: Point Cloud Transformer [paper]
• Transformer Interpretability Beyond Attention Visualization[paper] [code]
• [METRO] End-to-End Human Pose and Mesh Reconstruction with Transformers [paper]
• [PointTransformer] Point Transformer[paper]
• [PED] DETR for Pedestrian Detection[paper]
• [UP-DETR] UP-DETR: Unsupervised Pre-training for Object Detection with Transformers[paper]
• [LAMBDANETWORKS] MODELING LONG-RANGE INTERACTIONS WITHOUT ATTENTION[paper] [code]
• [C-Tran] General Multi-label Image Classification with Transformers[paper]
• [TSP-FCOS] Rethinking Transformer-based Set Prediction for Object Detection[paper]
• [IPT] Pre-Trained Image Processing Transformer[paper]
• [ACT] End-to-End Object Detection with Adaptive Clustering Transformer[paper]
• [VTs] Visual Transformers: Token-based Image Representation and Processing for Computer Vision[paper]

2021

• [Vision Transformer] An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale(ICLR)[paper] [code]
• [Deformable DETR] Deformable DETR: Deformable Transformers for End-to-End Object Detection(ICLR)[paper] [code]
• [LSTR] End-to-end Lane Shape Prediction with Transformers(WACV) [paper] [code]

2020

• [DETR] End-to-End Object Detection with Transformers (ECCV) [paper] [code]
• [FPT] Feature Pyramid Transformer(CVPR) [paper] [code]
• [TTSR] Learning Texture Transformer Network for Image Super-Resolution(CVPR) [paper] [code]

Reference

1. origin

Copyright

Collected by Lucas Jin. 2021