ashar6194/paper-summaries

Summaries for breakthrough research in contemporary deep learning

Paper-Summaries

Summaries for breakthrough research in contemporary deep learning. Arranged in chronological order

---

Image Recognition

1. Going Deeper with Convolutions (Inception Networks)
2. Deep Residual Learning for Image Recognition (ResNets)

Object Detection

1. Mask R-CNN
2. YOLO
3. SSD

More to come ---

Semantic Segmentation

1. Fully Convolutional Networks
2. Segnet
3. Deeplab

Optimization Frameworks

1. Mobilenets
2. SqueezeNets
3. ESPNets

3D Deep Learning

1. Learning Rich Features from RGB-D Images for Object Detection and Segmentation.
2. Frustum Pointnets
3. Voxelnets

Human Pose Estimation

2D-2D

1. Review
2. OpenPose

3D-3D

1. Real-Time Human Motion Capture with Multiple Depth Cameras

2D-3D

1. DensePose
2. A simple yet effective baseline for 3d human pose estimation

General Deep Learning Concepts

1. Update Methods for Network Optimization