Back to portfolio

Portfolio case study

Python Library: AutoToloka

Interactive segmentation toolkit that cuts labeling cost for CV datasets.

Overview

Python library that streams clicks/polygons to interactive models and pipelines the outputs to crowdsourcing or MLOps stacks.

What It Covers

  • Reduces labeling costs
  • Pipeline-friendly
  • Interactive masks

Stack And Topics

  • Python
  • Interactive Segmentation
  • Multi-modal Networks

Public Signals

  • Public references and qualitative delivery signals are used where numeric benchmarks are not public.

References