DeepPrior++: Improving Fast and Accurate 3D Hand Pose Estimation

Published in

(2017)

External Links:

Generate Graph DownloadPDF

TL;DR

With simple improvements: adding ResNet layers, data augmentation, and better initial hand localization, DeepPrior achieves better or similar performance than more sophisticated recent methods on the three main benchmarks (NYU, ICVL, MSRA) while keeping the simplicity of the original method.

Abstract

DeepPrior [18] is a simple approach based on Deep Learning that predicts the joint 3D locations of a hand given a depth map. Since its publication early 2015, it has been outperformed by several impressive works. Here we show that with simple improvements: adding ResNet layers, data augmentation, and better initial hand localization, we achieve better or similar performance than more sophisticated recent methods on the three main benchmarks (NYU, ICVL, MSRA) while keeping the simplicity of the original method. Our new implementation is available at https://github.com/moberweger/deep-prior-pp.

Authors

Markus Oberweger

3 papers

V. Lepetit

10 papers

References50 items

Correction to: Lie-X: Depth Image Based Articulated Object Pose Estimation, Tracking, and Action Recognition on Lie Groups

Multi-task, multi-domain learning: Application to semantic segmentation and pose regression

End-to-end Global to Local CNN Learning for Hand Pose Recovery in Depth data

Learning to Estimate 3D Hand Pose from Single RGB Images

Hand3D: Hand Pose Estimation using 3D Neural Network

DeepPrior++: Improving Fast and Accurate 3D Hand Pose Estimation

Published in

(2017)

External Links:

Generate Graph DownloadPDF

TL;DR

Abstract

Authors

Markus Oberweger

3 papers

V. Lepetit

10 papers

References50 items

Correction to: Lie-X: Depth Image Based Articulated Object Pose Estimation, Tracking, and Action Recognition on Lie Groups

Multi-task, multi-domain learning: Application to semantic segmentation and pose regression

End-to-end Global to Local CNN Learning for Hand Pose Recovery in Depth data

Learning to Estimate 3D Hand Pose from Single RGB Images

Hand3D: Hand Pose Estimation using 3D Neural Network

Crossing Nets: Dual Generative Models with a Shared Latent Space for Hand Pose Estimation

Region ensemble network: Improving convolutional network for hand pose estimation

Guided optimisation through classification and regression for hand pose estimation

Multivariate Regression with Grossly Corrupted Observations: A Robust Approach and its Applications

Hand Pose Regression via a Classification-Guided Approach

OctNet: Learning Deep 3D Representations at High Resolutions

Hand Pose Estimation from Local Surface Normals

Lie-X: Depth Image Based Articulated Object Pose Estimation, Tracking, and Action Recognition on Lie Groups

Efficient and precise interactive hand tracking through joint, continuous optimization of pose and correspondences

DeepHand: Robust Hand Pose Estimation by Completing a Matrix Imputed with Deep Features

Robust 3D Hand Pose Estimation in Single Depth Images: From Single-View CNN to Multi-View CNNs

Model-Based Deep Hand Pose Estimation

DISCO Nets : DISsimilarity COefficients Networks

Fits Like a Glove: Rapid and Reliable Hand Shape Personalization

Structured Prediction of 3D Human Pose with Deep Neural Networks

Spatial Attention Deep Net with Partial PSO for Hierarchical Hybrid Hand Pose Estimation

Deep Residual Learning for Image Recognition

Opening the Black Box: Hierarchical Sampling Optimization for Estimating Human Hand Pose

Training a Feedback Loop for Hand Pose Estimation

3D Hand Pose Estimation Using Randomized Decision Forest with Segmentation Index Points

Hand pose estimation through semi-supervised and weakly-supervised learning

A Framework for Articulated Hand Pose Estimation and Evaluation

Cascaded hand pose regression

Fast and robust hand tracking using detection-guided optimization

Depth-Based Hand Pose Estimation: Methods, Data, and Challenges

Hands Deep in Deep Learning for Hand Pose Estimation

Adam: A Method for Stochastic Optimization

Real-Time Continuous Pose Recovery of Human Hands Using Convolutional Networks

Going deeper with convolutions

Very Deep Convolutional Networks for Large-Scale Image Recognition

Latent Regression Forest: Structured Estimation of 3D Articulated Hand Posture

Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition

Realtime and Robust Hand Tracking from Depth

Real-Time Articulated Hand Pose Estimation Using Semi-supervised Transductive Regression Forests

Efficient Hand Pose Estimation from a Single Depth Image

Deep Convolutional Network Cascade for Facial Point Detection

Dynamics based 3D skeletal hand tracking

ImageNet classification with deep convolutional neural networks

Hand Pose Estimation and Hand Shape Classification Using Multi-layered Randomized Decision Forests

The Vitruvian manifold: Inferring dense correspondences for one-shot human pose estimation

Real time hand pose estimation using depth sensors

Full DOF tracking of a hand interacting with an object by modeling occlusions and physical constraints

Vision-based hand pose estimation: A review

Robust Articulatedicp for Realtime Hand Tracking

Dropout: a simple way to prevent neural networks from overfitting

Field of Study

Computer Science

Journal Information

Name

2017 IEEE International Conference on Computer Vision Workshops (ICCVW)

Page

1480-1489