3D Human Pose Lifting with Grid Convolution
Authors: Yangyuxuan Kang, Yuyang Liu, Anbang Yao, Shandong Wang, Enhua Wu
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on public 3D human pose estimation datasets demonstrate superior performance of our fully convolutional lifting network to existing methods by using the proposed representation learning paradigm. Furthermore, our method retains its effectiveness in the augmented training regime with synthetic data or joint optimization of the 2D human pose detector and the 2D-to-3D lifting network, showing further improved performance. |
| Researcher Affiliation | Collaboration | 1 State Key Laboratory of Computer Science, Institute of Software, Chinese Academy of Sciences 2 University of Chinese Academy of Sciences 3 Tsinghua University 4 Intel Labs China 5 Faculty of Science and Technology, University of Macau |
| Pseudocode | No | The paper does not contain a clearly labeled 'Pseudocode' or 'Algorithm' block. Figure 1 and 3 are diagrams, not pseudocode. |
| Open Source Code | Yes | Code is available at https://github.com/OSVAI/Grid Conv. |
| Open Datasets | Yes | Human3.6M. It is the largest indoor 3D human motion benchmark with 3D labels collected by motion capture system (Ionescu et al. 2014). |
| Dataset Splits | Yes | We follow the convention that takes Subject 1,5,6,7,8 as the training set and Subject 9,11 as the evaluation set. |
| Hardware Specification | Yes | We train and test the model on a single NVIDIA 1080Ti GPU. |
| Software Dependencies | No | The paper mentions using Adam optimizer but does not specify any software libraries or frameworks with version numbers (e.g., PyTorch version, Python version). |
| Experiment Setup | Yes | We train the model with Adam optimizer using a batch size of 200 and a learning rate starting at 0.001 for 100 epochs. |