# CF-DETR: A Lightweight Real-Time Model for Chicken Face Detection in High-Density Poultry Farming

## TL;DR

CF-DETR is a fast, accurate model for detecting chicken faces in crowded farms, helping improve poultry health monitoring and productivity.

## Contribution

CF-DETR introduces lightweight, real-time chicken face detection with improved accuracy and efficiency in dense environments.

## Key findings

- CF-DETR achieves 96.9% mean average precision at IoU 0.50 for chicken face detection.
- The model processes video at 81.4 frames per second with 33.2% fewer parameters than the baseline.
- Ablation studies confirm the effectiveness of each module in improving robustness and performance.

## Abstract

Modern poultry farms often use automated systems to monitor flock health, but overlapping chickens and changing lighting make individual face detection challenging. Traditional methods struggle to identify each bird reliably in these conditions. In this study, we present CF-DETR, a new deep-learning model that detects chicken faces in real time. The model uses special computer vision techniques to capture details even when birds overlap or move. We tested CF-DETR on images from crowded farm environments. It detected over 95% of chicken faces correctly and processed video at around 80 frames per second. These results show CF-DETR can robustly monitor chickens in dense flocks under challenging conditions. This automatic monitoring system could help farmers track poultry health earlier and more efficiently, improving animal welfare and farm productivity. By accurately detecting each chicken’s face, CF-DETR could help farmers spot signs of disease or stress earlier. Such smart monitoring is valuable for modern intelligent poultry farming.

Reliable individual detection under dense and cluttered conditions is a prerequisite for automated monitoring in modern poultry systems. We propose CF-DETR, an end-to-end detector that builds on RT-DETR and is tailored to chicken face detection in production-like environments. CF-DETR advances three technical directions: Dynamic Inception Depthwise Convolution (DIDC) expands directional and multi-scale receptive fields while remaining lightweight, Polar Embedded Multi-Scale Encoder (PEMD) restores global context and fuses multi-scale information to compensate for lost high-frequency details, and a Matchability Aware Loss (MAL) aligns predicted confidence with localization quality to accelerate convergence and improve discrimination. On a comprehensive broiler dataset, CF-DETR achieves a mean average precision at IoU 0.50 of 96.9% and a mean average precision (IoU 0.50–0.95) of 62.8%. Compared to the RT-DETR baseline, CF-DETR reduces trainable parameters by 33.2% and lowers FLOPs by 23.0% while achieving 81.4 frames per second. Ablation studies confirm that each module contributes to performance gains and that the combined design materially enhances robustness to occlusion and background clutter. Owing to its lightweight design, CF-DETR is well-suited for deployment in real-time smart farming monitoring systems. These results indicate that CF-DETR delivers an improved trade-off between detection performance and computational cost for real-time visual monitoring in intensive poultry production.