Scotopic Visual Recognition

Bo Chen, Pietro Perona; Proceedings of the IEEE International Conference on Computer Vision (ICCV) Workshops, 2015, pp. 8-11


Recognition from a small number of photons is important for biomedical imaging, security, astronomy and many other fields. We develop a framework that allows a machine to classify objects as quickly as possible, hence requiring as few photons as possible, while maintaining the error rate below an acceptable threshold. The framework also allows for a dynamic speed versus accuracy tradeoff. Given a generative model of the scene, the optimal tradeoff can be obtained from operations akin to the feedforward computation in a deep neural network. The generative model may also be learned from the data. The learned model requires less than 1 photon per pixel to achieve the same performance obtained with images in normal lighting conditions on the MNIST dataset, and 10 photons per unit to be within 1% of the best accuracy of the CIFAR10 dataset.

Related Material

author = {Chen, Bo and Perona, Pietro},
title = {Scotopic Visual Recognition},
booktitle = {Proceedings of the IEEE International Conference on Computer Vision (ICCV) Workshops},
month = {December},
year = {2015}