Task-Specific Zero-shot Quantization-Aware Training for Object Detection
This work addresses the challenge of quantizing object detection models without access to real training data, which is important for deployment in privacy-sensitive or resource-constrained environments, though it is incremental over existing zero-shot quantization methods.
The paper tackles the problem of zero-shot quantization for object detection networks by proposing a task-specific framework that synthesizes calibration data with bounding box and category information, achieving state-of-the-art performance on MS-COCO and Pascal VOC datasets.
Quantization is a key technique to reduce network size and computational complexity by representing the network parameters with a lower precision. Traditional quantization methods rely on access to original training data, which is often restricted due to privacy concerns or security challenges. Zero-shot Quantization (ZSQ) addresses this by using synthetic data generated from pre-trained models, eliminating the need for real training data. Recently, ZSQ has been extended to object detection. However, existing methods use unlabeled task-agnostic synthetic images that lack the specific information required for object detection, leading to suboptimal performance. In this paper, we propose a novel task-specific ZSQ framework for object detection networks, which consists of two main stages. First, we introduce a bounding box and category sampling strategy to synthesize a task-specific calibration set from the pre-trained network, reconstructing object locations, sizes, and category distributions without any prior knowledge. Second, we integrate task-specific training into the knowledge distillation process to restore the performance of quantized detection networks. Extensive experiments conducted on the MS-COCO and Pascal VOC datasets demonstrate the efficiency and state-of-the-art performance of our method. Our code is publicly available at: https://github.com/DFQ-Dojo/dfq-toolkit .