Simple Baselines for Human Pose Estimation and Tracking

By Bin Xiao, Haiping Wu, Yichen Wei. 2018

2,084 citations recorded by the source corpus.

There has been significant progress on pose estimation and increasing interests on pose tracking in recent years. At the same time, the overall algorithm and system complexity increases as well, making the algorithm analysis and comparison more difficult. This work provides simple and effective baseline methods. They are helpful for inspiring and evaluating new ideas for the field. State-of-the-art results are achieved on challenging benchmarks. The code will be available at this https URL.

Subfields: Computer Vision and Pattern Recognition, Human-Computer Interaction

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Human-Computer Interaction

Citations in the corpus

Objects as Points (2019, methodology): “With a simple Resnet-18 and up-convolutional layers [55], our network runs at 142 FPS with 28 .”
Grounded Language-Image Pre-training (2021, background): “However, to gain fine-grained understanding of images, as required by many tasks, such as object detection [33, 46], segmentation [7,37], human pose estimation [51,58], scene understanding [15, 27, 59], action recognition [19], visionlanguage understanding [8,29–32,38,50,52,65,67], objectlevel visua”
Incremental Few-Shot Object Detection (2020, positive): “In practice, the feature extractor model is implemented with the ResNet-based backbone of [52].”
Deep High-Resolution Representation Learning for Visual Recognition (2019, methodology): “(iii) Compared to the previous best-performed method SimpleBaseline [152], our HRNetV 1 -”
ResNeSt: Split-Attention Networks (2020, background): “Networks trained for image classification often serve as the backbone of the neural networks designed for other applications, such as object detection [22,46], semantic segmentation [6, 43, 73] and pose estimation [14, 58].”
Multi-Scale Vision Longformer: A New Vision Transformer for High-Resolution Image Encoding (2021, background): “But its quadratic increase of both computational and memory complexity hinders its application on many vision tasks that require highresolution feature maps computed on high-resolution images1, like object detection [31, 23], segmentation [25, 5], and human pose estimation [45, 34].”
Deep High-Resolution Representation Learning for Human Pose Estimation (2019, methodology): “video pose estimation and tracking [48, 70], etc.”
Bottom-Up Object Detection by Grouping Extreme and Center Points (2019, methodology): “The network is trained in a fully supervised way, with either an L2 loss to a rendered Gaussian map [3,5,30,49] or with a per-pixel logistic regression loss [22, 34, 35].”
Res2Net: A New Multi-Scale Backbone Architecture (2019, methodology): “.5 81.4 70.8 78.2 ResNet-101 71.4 89.3 79.3 68.1 78.1 Res2Net-101 74.4 92.6 82.6 72.0 78.5 using the COCO validation set. Following common settings, we use the same person detectors in SimpleBaseline [55] for evaluation. Table10shows the performance of key-points estimation on the COCO validation se”
ViTAEv2: Vision Transformer Advanced by Exploring Inductive Bias for Image Recognition and Beyond (2022, methodology): “Following the setting in AP-10K, we adopt SimpleBaseline [82] as the pose estimation framework and train the models with various backbones for 210 epochs using Adam optimizer and images of size 256× 256.”
Probabilistic Regression for Visual Tracking (2020, background): “Along with a myriad of other computer vision tasks, such as object detection [24, 26, 35], pose estimation [6, 42, 48], and keypoint detection [36, 46], visual tracking can fundamentally be formulated as a regression problem.”
Effective Whole-body Pose Estimation with Two-stages Distillation (2023, methodology): “Due to the lack of whole-body annotations, most previous models are designed for body-only [22, 40, 47, 53], hand-only [6, 32, 62, 68], or face-only [21, 46, 67].”
SLEAP: A deep learning system for multi-animal pose tracking (2022, methodology): “First, we implemented a flow-shift-based tracking approach 23 that uses optical flow to estimate the displacement of poses across frames.”
Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose (2018, background): “The skeleton consists of keypoints (or joints): ankles, knees, hips, elbows, etc. Human pose estimation accuracy was greatly improved with the help of convolutional neural networks (CNNs) [6], [4], [16].”
Learnable Triangulation of Human Pose (2019, methodology): “Then we feed the cropped images I c into a deep convolutional neural network backbone based on the ”simple baselines” architecture [21].”
ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation (2022, methodology): “The detection results from SimpleBaseline [44] are utilized for evaluating ViTPose’s performance on the MS COCO Keypoint val set.”