Sector Patch Embedding: An Embedding Module Conforming to The Distortion Pattern of Fisheye Image
This work addresses a domain-specific problem for computer vision applications using fisheye cameras, offering an incremental improvement by optimizing Transformer-based models for distortion perception.
The paper tackles the problem of poor performance in fisheye vision tasks due to image distortion by proposing Sector Patch Embedding (SPE), a novel patch embedding method that conforms to distortion patterns, resulting in improved classification top-1 accuracy by 0.75% for ViT and 2.8% for PVT on a synthetic fisheye dataset.
Fisheye cameras suffer from image distortion while having a large field of view(LFOV). And this fact leads to poor performance on some fisheye vision tasks. One of the solutions is to optimize the current vision algorithm for fisheye images. However, most of the CNN-based methods and the Transformer-based methods lack the capability of leveraging distortion information efficiently. In this work, we propose a novel patch embedding method called Sector Patch Embedding(SPE), conforming to the distortion pattern of the fisheye image. Furthermore, we put forward a synthetic fisheye dataset based on the ImageNet-1K and explore the performance of several Transformer models on the dataset. The classification top-1 accuracy of ViT and PVT is improved by 0.75% and 2.8% with SPE respectively. The experiments show that the proposed sector patch embedding method can better perceive distortion and extract features on the fisheye images. Our method can be easily adopted to other Transformer-based models. Source code is at https://github.com/IN2-ViAUn/Sector-Patch-Embedding.