Papers of the day   All papers

End-to-End Object Detection with Transformers

Comments

Andrej Karpathy: DE⫶TR: End-to-End Object Detection with Transformers https://github.com/facebookresearch/detr (+paper https://arxiv.org/abs/2005.12872) awesome to see a solid swing at (non-autoregressive) end-to-end detection. Anchor boxes + nms is a mess. (I was hoping detection would go end-to-end back in ~2013) https://t.co/cN4hK2ZFY0

16 replies, 1500 likes


roadrunner01: End-to-End Object Detection with Transformers pdf: https://arxiv.org/pdf/2005.12872.pdf abs: https://arxiv.org/abs/2005.12872 github: https://github.com/facebookresearch/detr https://t.co/Kxnpptn6Z3

1 replies, 321 likes


phalanx: End-to-End Object Detection with Transformers arxiv: https://arxiv.org/abs/2005.12872 code: https://github.com/facebookresearch/detr https://t.co/ckjtw19n5H

0 replies, 121 likes


arXiv CS-CV: End-to-End Object Detection with Transformers http://arxiv.org/abs/2005.12872

0 replies, 11 likes


Daisuke Okanohara: Object detection can be treated as a set prediction problem. DETR replaces complex post-processing with transformer s.t. an encoder aggregates global information, and a decoder outputs a set of proposals, and measure the loss via bipartite matching. https://arxiv.org/abs/2005.12872

0 replies, 5 likes


Nikolai Yakovenko: Finally finished the paper, it’s perhaps a bit long. But great work, great way to simplify what has been a messy problem. The kind of paper that reminds you what our field is all about. Highly recommended — even for people relatively new to the space.

1 replies, 4 likes


Akshaj Verma: End-to-End Object Detection with Transformers by Facebook AI DETR uses a conventional CNN backbone to learn a 2D representation of an input image. #deeplearning #machinelearning #PyTorch #research https://arxiv.org/pdf/2005.12872.pdf https://t.co/zzda4lWPe4

0 replies, 4 likes


HotComputerScience: Most popular computer science paper of the day: "End-to-End Object Detection with Transformers" https://hotcomputerscience.com/paper/end-to-end-object-detection-with-transformers https://twitter.com/karpathy/status/1265742371649544192

0 replies, 2 likes


Jovian.ml: DE⫶TR: End-to-End Object Detection with Transformers PyTorch training code and pretrained models for DETR (DEtection TRansformer). Paper: https://arxiv.org/abs/2005.12872 Code: https://github.com/facebookresearch/detr #datascience #100daysofcode #deeplearning #PyTorch

0 replies, 2 likes


arXiv CS-CV: End-to-End Object Detection with Transformers http://arxiv.org/abs/2005.12872

0 replies, 1 likes


cs.CV Papers: End-to-End Object Detection with Transformers. Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, and Sergey Zagoruyko http://arxiv.org/abs/2005.12872

1 replies, 0 likes


Content

Found on May 27 2020 at https://arxiv.org/pdf/2005.12872.pdf

PDF content of a computer science paper: End-to-End Object Detection with Transformers