DiffYOLO: Object Detection for Anti-Noise via YOLO and Diffusion Models
This addresses the challenge of object detection in non-ideal conditions for users relying on YOLO models, though it appears incremental as it builds on existing YOLO and diffusion methods.
The authors tackled the problem of object detection on low-quality noisy datasets by proposing DiffYOLO, a framework that enhances well-trained YOLO models using feature maps from denoising diffusion probabilistic models, allowing fine-tuning on high-quality data and testing on low-quality data, with results showing improved performance on both noisy and high-quality datasets.
Object detection models represented by YOLO series have been widely used and have achieved great results on the high quality datasets, but not all the working conditions are ideal. To settle down the problem of locating targets on low quality datasets, the existing methods either train a new object detection network, or need a large collection of low-quality datasets to train. However, we propose a framework in this paper and apply it on the YOLO models called DiffYOLO. Specifically, we extract feature maps from the denoising diffusion probabilistic models to enhance the well-trained models, which allows us fine-tune YOLO on high-quality datasets and test on low-quality datasets. The results proved this framework can not only prove the performance on noisy datasets, but also prove the detection results on high-quality test datasets. We will supplement more experiments later (with various datasets and network architectures).