面向目标检测的输入退化量化鲁棒性

摘要

后训练量化（PTQ）对于在资源受限设备上部署高效的目标检测模型（如YOLO）至关重要。然而，降低精度对模型在面对现实世界输入退化（如噪声、模糊和压缩伪影）时的鲁棒性影响是一个重要问题。本文通过一项全面的实证研究，评估了YOLO模型（从nano到extra-large规模）在多种精度格式下的鲁棒性：FP32、FP16（TensorRT）、动态UINT8（ONNX）和静态INT8（TensorRT）。我们提出并评估了一种针对静态INT8 PTQ的退化感知校准策略，其中TensorRT校准过程暴露于混合了干净图像和合成退化图像的样本中。模型在COCO数据集上进行了基准测试，涵盖了七种不同的退化条件（包括各种类型和程度的噪声、模糊、低对比度及JPEG压缩）以及一种混合退化场景。结果表明，尽管静态INT8 TensorRT引擎在干净数据上提供了显著的加速（约1.5-3.3倍）且准确率下降适中（约3-7% mAP50-95），但所提出的退化感知校准在大多数模型和退化情况下并未带来一致且广泛的鲁棒性提升，相较于标准的干净数据校准。一个显著的例外是在特定噪声条件下，较大规模的模型表现出了改进，这表明模型容量可能影响此校准方法的有效性。这些发现凸显了增强PTQ鲁棒性的挑战，并为在非受控环境中部署量化检测器提供了洞见。所有代码和评估表格均可在https://github.com/AllanK24/QRID获取。

English

Post-training quantization (PTQ) is crucial for deploying efficient object detection models, like YOLO, on resource-constrained devices. However, the impact of reduced precision on model robustness to real-world input degradations such as noise, blur, and compression artifacts is a significant concern. This paper presents a comprehensive empirical study evaluating the robustness of YOLO models (nano to extra-large scales) across multiple precision formats: FP32, FP16 (TensorRT), Dynamic UINT8 (ONNX), and Static INT8 (TensorRT). We introduce and evaluate a degradation-aware calibration strategy for Static INT8 PTQ, where the TensorRT calibration process is exposed to a mix of clean and synthetically degraded images. Models were benchmarked on the COCO dataset under seven distinct degradation conditions (including various types and levels of noise, blur, low contrast, and JPEG compression) and a mixed-degradation scenario. Results indicate that while Static INT8 TensorRT engines offer substantial speedups (~1.5-3.3x) with a moderate accuracy drop (~3-7% mAP50-95) on clean data, the proposed degradation-aware calibration did not yield consistent, broad improvements in robustness over standard clean-data calibration across most models and degradations. A notable exception was observed for larger model scales under specific noise conditions, suggesting model capacity may influence the efficacy of this calibration approach. These findings highlight the challenges in enhancing PTQ robustness and provide insights for deploying quantized detectors in uncontrolled environments. All code and evaluation tables are available at https://github.com/AllanK24/QRID.

面向目标检测的输入退化量化鲁棒性

Quantization Robustness to Input Degradations for Object Detection

摘要

Support