Looking for a Researcher/Developer to Reproduce a Deep Learning Paper
I need help reproducing the results of a published computer vision and human pose estimation paper. The work involves 3D human pose estimation from video using graph-based transformer architectures.
What you'll do:
- Set up and run the provided codebase (PyTorch-based).
- Reproduce training and evaluation on standard benchmark datasets.
- Match the reported performance metrics from the paper.
Required skills:
- Strong experience with PyTorch and deep learning.
- Familiarity with human pose estimation or skeleton-based action recognition.
- Experience with graph neural networks (GNN) and/or transformers.
- Comfortable working with research code and HPC/GPU environments.
- Ability to read and implement from academic papers.
Please include examples of similar work (paper reproduction, research implementation, or pose estimation projects) in your proposal.
Apply tot his job
Apply To this Job