sellCollaboration Call sellContinual Learning sellHuman Pose Estimation sellModel Optimization

Call: Continual Adaptation and Optimization for Pose Models

Muhammad Saif Ullah Khan

Abstract

This call targets continual adaptation of human pose models under domain and task shift, with emphasis on resource constraints, sparsity, and reliable long-horizon evaluation.

Scope

This collaboration area studies how to keep pose models useful over time when domains, sensors, keypoint definitions, and deployment constraints change.

A central objective is to move beyond one-off retraining toward adaptation strategies that remain efficient, stable, and reproducible.

Open Collaboration Tasks

  1. Continual adaptation benchmarks for domain and class/keypoint increments in pose estimation.
  2. Memory- and compute-aware adaptation under strict update budgets.
  3. Sparsity, pruning, and compact adaptation modules for sustainable deployment.
  4. Evaluation of forgetting, transfer, and stability across multi-step adaptation sequences.

What Is Already Available

  • PoseAdapt benchmark structure and initial method baselines.
  • Existing pose modeling pipelines that can serve as adaptation backbones.
  • Reproducibility-oriented training/evaluation workflow from earlier projects.

Collaboration Format

  • Joint experimental design, protocol definition, and baselining.
  • Shared implementation with transparent reporting of constraints and tradeoffs.
  • Co-authored papers and open benchmark artifacts when results mature.

To start a discussion, send a short collaboration note to mukh07@dfki.de including your relevant background and the adaptation setting you want to target.

Maintained by saifkhichi96 on GitHub.

The website is distributed under different open-source licenses. For more details, see the notice at the bottom of the page.