Adapting CNNs for Fisheye Cameras without Retraining
This work addresses the challenge of using fisheye cameras in computer vision applications without requiring retraining, which is incremental as it builds on existing CNN architectures.
The paper tackles the problem of adapting pre-trained convolutional networks to operate on fisheye camera images without retraining, by proposing Rectified Convolutions (RectConv), which enables networks to handle both rectified patches and the full field of view, and demonstrates its effectiveness on segmentation and detection tasks using two public datasets.
The majority of image processing approaches assume images are in or can be rectified to a perspective projection. However, in many applications it is beneficial to use non conventional cameras, such as fisheye cameras, that have a larger field of view (FOV). The issue arises that these large-FOV images can't be rectified to a perspective projection without significant cropping of the original image. To address this issue we propose Rectified Convolutions (RectConv); a new approach for adapting pre-trained convolutional networks to operate with new non-perspective images, without any retraining. Replacing the convolutional layers of the network with RectConv layers allows the network to see both rectified patches and the entire FOV. We demonstrate RectConv adapting multiple pre-trained networks to perform segmentation and detection on fisheye imagery from two publicly available datasets. Our approach requires no additional data or training, and operates directly on the native image as captured from the camera. We believe this work is a step toward adapting the vast resources available for perspective images to operate across a broad range of camera geometries.