Computer Vision Paper Reading for ISCAS
- 科研在线团队文档库
- 注册科研在线帐号
- 点击加入ISCAS007链接 似乎此处附带的注册功能失效了！！！但点击加入估计还行
- ccf 推荐排名
- ai deadline
- github deadline
- wiki call for paper
实测：163邮箱，教育网邮箱，QQ邮箱可以注册，国科大邮箱可以直接登录。 如果点击注册没有反应，建议使用最新版本chrome浏览器。
| A | 朱玉影 | 杜肖冰 | 窦毅琨 | 李伟亨 |
|---|---|---|---|---|
| B | 刘舫 | 张拯明 | 晁文涛 | 康杨雨轩 |
| C | 王佳欣 | 石玥 | 曲文天 | 李锦瑶 |
| D | 林泽一 | 吴通通 | 朱倩 | 左德鑫 |
| E | 陈紫檀 | 薛涵 | 左然 | 宋建成 |
-
Deep Reinforcement Learning for Unsupervised Video Summarization 朱倩
-
Point-to-Pose Voting based Hand Pose Estimation using Residual Permutation Equivariant Layer 左德鑫
-
MURAUER: Mapping Unlabeled Real Data for Label AUstERity 左德鑫
Abstract: the author present an Adaptive Octree-based Convolutional Neural Network (Adaptive O-CNN) for efficient 3D shape encoding and decoding
- 窦毅琨
- 杜肖冰
-
[Bring it to the Pitch: Combining Video and Movement Data to Enhance Team Sport Analysis-TVCG2018] 朱倩 pdf ppt
- [Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics] 王佳欣 pdf ppt author
- A survey of human pose estimation of cvpr2017 ppt 朱玉影 pdf1 - Cascaded Pyramid Network for Multi-Person Pose Estimation pdf2 - Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose pdf3 - A simple yet effective baseline for 3d human pose estimation
Abstract: limited by dataset,we can do human pose detection problem to get 2d heatmap or do regression problem to get 3d joints coordinate.So in this survey, I introduce three paper solve human pose estimation problem in three ways,i)rgb->2d heatmap(2d joints),ii)rgb->3d joints(3d heatmap),iii)2d joints->3d joints.
- 草图主题论文831 薛涵
Abstract: we can solve hand pose estimation problem as regression problem which regress joints location(xyz coordinates) or classification(detection/localization) problem which show its probility in input image or voxels for each joint. So we can combine two solution and introduce in dense guidance map as intermediate predictions. So this paper design seveal guidance map and show its addvantage in experiments.
- Stacked Hourglass Networks 朱玉影
Learning to Estimate 3D Human Pose and Shape from a Single Color Image PPT/PDF Stacked Hourglass Networks for Human Pose Estimation PDF/tensorflow_code
- 郑文勇
- ECGLens
左然朱倩（临时调换） - 左德鑫
- Context Encoding for Semantic Segmentation 王佳欣 Paper Code Team
- 林泽一
- 吴通通
- 3D point 窦毅琨
- BodyNet: Volumetric Inference of 3D Human Body Shapes project page/pdf/ppt 朱玉影
Abstract: BodyNet is an end-to-end trainable network for human shape estimation that benefits from (i) a volumetric 3D loss, (ii) a multi-view re-projection loss, and (iii) intermediate supervision of 2D pose, 2D body part segmentation, and 3D pose.
- Instance Segmentation 王佳欣
- pose 郑文勇
- Semantic Segmentation 王佳欣
- Deformable GANs for Pose-based Human Image Generation 曲文天 slides
- Deformable GANs for Pose-based Human Image Generation 曲文天 pdf
- Geometric deep learning on graphs and manifolds using mixture model CNNs 朱玉影 slides
- Geometric deep learning on graphs and manifolds using mixture model CNNs 朱玉影 pdf
- DracNets pptx 王佳欣 slides
- DiracNets: Training Very Deep Neural Networks Without Skip-Connections 2018
- DenseNet 王佳欣 paper
- DenseNet slides DenseNet 原作者 CVPR slides
- DenseNet 王佳欣 slides
- Hybrid VAE, Improving Deep Generative Models using Partial Observations 郑文勇 paper
- Hybrid VAE, Improving Deep Generative Models using Partial Observations 郑文勇 ppt
- ResNet 王佳欣 paper
- Highway Networks 王佳欣 paper
- Recent Advances in Convolutional Neural Networks 王佳欣 paper
- Recent Advances in Convolutional Neural Networks 王佳欣 slides
- End-to-end Recovery of Human Shape and Pose 朱玉影 ppt
- End-to-end Recovery of Human Shape and Pose 朱玉影 pdf
- AI and Deep Learning in 2017 – A Year in Review
- Deep learning Theoies From Empirical to Theory
- 2017 paper reading 记录