arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2508.19600cs.CV
收藏

量化对输入退化鲁棒性的研究:目标检测

Quantization Robustness to Input Degradations for Object Detection

Bahcesehir University(巴赫塞希尔大学)

Toghrul Karimov, Hassan Imani, Allan Kazakov

AI总结:

本文研究了量化对目标检测模型在真实输入退化下的鲁棒性影响,通过评估YOLO模型在不同精度下的表现,发现静态INT8量化在清洁数据上速度提升显著,但退化感知校准在多数情况下未提升鲁棒性,但大模型在特定噪声下有例外。

摘要:

后训练量化(PTQ)对于在资源受限设备上部署高效目标检测模型(如YOLO)至关重要。然而,降精度对模型在现实输入退化(如噪声、模糊和压缩伪影)下的鲁棒性影响是一个重要问题。本文提出了一个全面的实证研究,评估了YOLO模型(从nano到extra-large规模)在多种精度格式(FP32、FP16(TensorRT)、动态UINT8(ONNX)和静态INT8(TensorRT))下的鲁棒性。我们引入并评估了一种退化感知的校准策略用于静态INT8 PTQ,其中TensorRT校准过程会接触到清洁和合成退化的图像混合。模型在COCO数据集上在七种不同的退化条件下(包括各种类型和级别的噪声、模糊、低对比度和JPEG压缩)以及混合退化场景下进行了基准测试。结果表明,尽管静态INT8 TensorRT引擎在清洁数据上提供显著的速度提升(约1.5-3.3倍),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的鲁棒性提升。值得注意的是,在特定噪声条件下,较大模型规模表现出例外,表明模型容量可能影响该校准方法的有效性。这些发现突显了增强PTQ鲁棒性的挑战,并为在不受控环境中部署量化检测器提供了见解。所有代码和评估表格均可在https://github.com/AllanK24/QRID上获得。

英文摘要:

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.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑