# Chanakya: Learning Runtime Decisions for Adaptive Real-Time Perception

**NeurIPS 2023**

*Advances in Neural Information Processing Systems (NeurIPS)*

Anurag Ghosh, **Vaibhav Balloli**, Akshay Nambi, Aditya Singh, Tanuja Ganu

## Summary

A learned approximate-execution framework for streaming perception. Chanakya jointly considers scene content and system contention to pick run-time decisions (resolution, model, compute) that balance accuracy and latency, beating static and dynamic baselines on both server GPUs and edge devices.

## Identifiers

- DOI: [10.52202/075280-2429](https://doi.org/10.52202/075280-2429)
- PDF: https://openreview.net/pdf?id=VpCjozUOM2

## Links

- Paper: https://openreview.net/pdf?id=VpCjozUOM2
- Code: https://github.com/microsoft/Chanakya
- Project: https://aka.ms/chanakya

## Topics

vision

## Citation

```bibtex
@inproceedings{ghosh2023chanakya,
  title     = {Chanakya: Learning Runtime Decisions for Adaptive Real-Time Perception},
  author    = {Ghosh, Anurag and Balloli, Vaibhav and Nambi, Akshay and Singh, Aditya and Ganu, Tanuja},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year      = {2023}
}
```

---

Canonical HTML page: https://profile.vballoli.com/papers/chanakya/