EfficientHRNet: Efficient Scaling for Lightweight High-Resolution Multi-Person Pose Estimation
arXiv:2007.08090
Abstract
There is an increasing demand for lightweight multi-person pose estimation for many emerging smart IoT applications. However, the existing algorithms tend to have large model sizes and intense computational requirements, making them ill-suited for real-time applications and deployment on resource-constrained hardware. Lightweight and real-time approaches are exceedingly rare and come at the cost of inferior accuracy. In this paper, we present EfficientHRNet, a family of lightweight multi-person human pose estimators that are able to perform in real-time on resource-constrained devices. By unifying recent advances in model scaling with high-resolution feature representations, EfficientHRNet creates highly accurate models while reducing computation enough to achieve real-time performance. The largest model is able to come within 4.4% accuracy of the current state-of-the-art, while having 1/3 the model size and 1/6 the computation, achieving 23 FPS on Nvidia Jetson Xavier. Compared to the top real-time approach, EfficientHRNet increases accuracy by 22% while achieving similar FPS with 1/3 the power. At every level, EfficientHRNet proves to be more computationally efficient than other bottom-up 2D human pose estimation approaches, while achieving highly competitive accuracy.
11 pages (13 with references), 3 figures
References in corpus (31)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
- An overview of gradient descent optimization algorithms
- Deep Residual Learning for Image Recognition
- DeepPose: Human Pose Estimation via Deep Neural Networks
- DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
- High-Resolution Representations for Labeling Pixels and Regions
- Learning Deconvolution Network for Semantic Segmentation
- Associative Embedding: End-to-End Learning for Joint Detection and Grouping
- Human pose estimation via Convolutional Part Heatmap Regression
- Stacked Hourglass Networks for Human Pose Estimation
- Deep High-Resolution Representation Learning for Visual Recognition
- Pyramid Scene Parsing Network
- Convolutional Pose Machines
- UNet++: A Nested U-Net Architecture for Medical Image Segmentation
- Learning Human Pose Estimation Features with Convolutional Networks
- Convolutional Neural Fabrics
- Attention to Scale: Scale-aware Semantic Image Segmentation
- Cascaded Pyramid Network for Multi-Person Pose Estimation
- DeeperCut: A Deeper, Stronger, and Faster Multi-Person Pose Estimation Model
- Multi-Context Attention for Human Pose Estimation
- Stacked Deconvolutional Network for Semantic Segmentation
- HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
- Learning Feature Pyramids for Human Pose Estimation
- Interlinked Convolutional Neural Networks for Face Parsing
- ExFuse: Enhancing Feature Fusion for Semantic Segmentation
- Decoupled Classification Refinement: Hard False Positive Suppression for Object Detection
- DeepCut: Joint Subset Partition and Labeling for Multi Person Pose Estimation
- Multi-Person Pose Estimation with Local Joint-to-Person Associations
- Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes