CV Code | 计算机视觉开源周报20191002期

图片来自ECA-Net

总结了过去一周新出的计算机视觉开源代码。

涵盖的方向包括域适应、消除偏差网络训练、视觉注意力模型、机器人动作搜索、机器人抓取、自动驾驶、神经架构搜索、文本识别、视觉搜索等。

域适应

MixMatch Domain Adaptaion: Prize-winning solution for both tracks of VisDA 2019 challenge

Danila Rukhovich, Danil Galeev

TASK-CV 2019 at ICCV

https://arxiv.org/abs/1910.03903v1

https://github.com/filaPro/visda2019

网络训练消除偏差影响 | 基于对抗性训练策略的方法来学习无偏且不变的区分特征

Bias-Resilient Neural Network

Ehsan Adeli, Qingyu Zhao, Adolf Pfefferbaum, Edith V. Sullivan, Li Fei-Fei, Juan Carlos Niebles, Kilian M. Pohl

https://arxiv.org/abs/1910.03676v1

http://github.com/QingyuZhao/BR-Net/

极轻量级的通道注意力模块，改进深度CNN网络

ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks

Qilong Wang, Banggu Wu, Pengfei Zhu, Peihua Li, Wangmeng Zuo, Qinghua Hu

https://arxiv.org/abs/1910.03151v1

https://github.com/BangguWu/ECANet

多源域适应 + 半监督域适应

Multi-Source Domain Adaptation and Semi-Supervised Domain Adaptation with Focus on Visual Domain Adaptation Challenge 2019

Yingwei Pan, Yehao Li, Qi Cai, Yang Chen, Ting Yao

https://arxiv.org/abs/1910.03548v1

https://github.com/Panda-Peter/visda2019-multisource

https://github.com/Panda-Peter/visda2019-semisupervised

机器人动作搜索规划

Object-centric Forward Modeling for Model Predictive Control

Yufei Ye, Dhiraj Gandhi, Abhinav Gupta, Shubham Tulsiani

https://arxiv.org/abs/1910.03568v1

https://judyye.github.io/ocmpc/

用视觉计算辅助互联网仇恨语音的检测

Hate Speech in Pixels: Detection of Offensive Memes towards Automatic Moderation

Benet Oriol Sabat, Cristian Canton Ferrer, Xavier Giro-i-Nieto

AI for Social Good Workshop at NeurIPS 2019

https://arxiv.org/abs/1910.02334v1

https://github.com/imatge-upc/hate-speech-detection

机器人抓去透明物体的3D形状估计

ClearGrasp: 3D Shape Estimation of Transparent Objects for Manipulation

Shreeyak S. Sajjan (1), Matthew Moore (1), Mike Pan (1), Ganesh Nagaraja (1), Johnny Lee (2), Andy Zeng (2), Shuran Song (2 and 3) ((1) Synthesis.ai (2) Google (3) Columbia University)

https://arxiv.org/abs/1910.02550v1

https://sites.google.com/view/cleargrasp

自动驾驶，长程视觉与语音导航

Talk2Nav: Long-Range Vision-and-Language Navigation in Cities

Arun Balajee Vasudevan, Dengxin Dai, Luc Van Gool

https://arxiv.org/abs/1910.02029v1

https://www.trace.ethz.ch/publications/2019/talk2nav/index.html

4小时GPU时间的鲁棒神经架构搜索

Searching for A Robust Neural Architecture in Four GPU Hours

Xuanyi Dong, Yi Yang

https://arxiv.org/abs/1910.04465v1

https://github.com/D-X-Y/NAS-Projects

部分遮挡物体的快速点云配准算法AlignNet-3D

AlignNet-3D: Fast Point Cloud Registration of Partially Observed Objects

Johannes Groß, Aljosa Osep, Bastian Leibe

3DV'19

https://arxiv.org/abs/1910.04668v1

https://www.vision.rwth-aachen.de/page/alignnet

任意形状文本识别，基于2D自注意力机制

On Recognizing Texts of Arbitrary Shapes with 2D Self-Attention

Junyeop Lee, Sungrae Park, Jeonghun Baek, Seong Joon Oh, Seonghyeon Kim, Hwalsuk Lee

https://arxiv.org/abs/1910.04396v1

（代码将开源，还未公布地址）

京东开源的实时视觉搜索系统

The Design and Implementation of a Real Time Visual Search System on JD E-commerce Platform

Jie Li, Haifeng Liu, Chuanghua Gui, Jianyu Chen, Zhenyun Ni, Ning Wang

https://arxiv.org/abs/1908.07389

https://github.com/vearch/vearch