Benchmarking ResNet Backbones in RT-DETR: Impact of Depth and Regularization under environmental conditions

TL;DR

This study evaluates how different ResNet backbones affect real-time object detection performance under environmental variations in robotics, highlighting the importance of backbone choice based on specific environmental challenges.

Contribution

It provides a comparative analysis of ResNet18, 34, 50, and 101 in RT-DETR, revealing how depth and regularization influence detection accuracy, confidence, and latency under environmental changes.

Findings

ResNet50 offers the best trade-off under illumination variation.

ResNet34 provides balanced performance under background variation.

Environmental conditions mainly affect confidence, not latency or accuracy.

Abstract

Visual perception plays a central role in competitive robotics, where environmental variations can directly affect real-time detection performance. The related literature on transformer-based detectors lack information regarding the impact of backbone scale and environmental settings on model performance. This work presents a comparative evaluation of RT-DETR for detecting round objects under environmental and hyperparameter variations relevant to competitive robotics. Four ResNet backbones (ResNet18, ResNet34, ResNet50, and ResNet101) were compared using dropout rates, analyzing their effect on confidence and accuracy. All models were trained under the same configuration and evaluated under changes in lighting and background contrast. Environmental conditions primarily impact prediction confidence, while inference latency remains largely unaffected and classification accuracy stays consistently high, approaching or above 1.00 in most cases. Two distinct behaviors were observed. Under illumination variation, ResNet50 achieves the best trade-off, combining near-perfect accuracy, confidence values up to approximately 0.869 and latency around 0.058-0.059 ms. Under background variation, ResNet34 provides the most balanced performance, reaching near-perfect accuracy and higher confidence values up to approximately 0.887. These results indicate that the optimal architecture depends on the type of environmental variation, with intermediate-depth models offering the best balance between performance and efficiency.