---
license: apache-2.0
datasets:
- detection-datasets/coco
pipeline_tag: object-detection
---

# Introduction

This repository stores the model for Faster-RCNN (ResNext50), compatible with Kalray's neural network API. </br>
Please see www.github.com/kalray/kann-model-zoo for details and proper usage [WIKI](https://github.com/kalray/kann-model-zoo/wiki). </br>

# Contents

- ONNX: retinanet-resnext50-fpn-800.onnx

# Lecture note reference

- Focal Loss for Dense Object Detection, https://arxiv.org/abs/1708.02002

# Repository or links references

- [Link to Download](https://zenodo.org/record/6617879/files/resnext50_32x4d_fpn.onnx)

  ```bibtex
  Kiswani, A., & Gonzalez, P. (2022).
  Trained Model for retinanet_resnext50_32x4d_fpn ONNX for MLPerf Inference (MLPerf v2.1 inference).
  Zenodo. https://doi.org/10.5281/zenodo.6617879
  ```

Authors:
+ qmuller@kalrayinc.com