Classification of optics-free images with deep neural networks

TL;DR

This paper demonstrates that deep neural networks can accurately classify optics-free images directly from sensor data, enabling privacy-preserving and power-efficient imaging without traditional optics.

Contribution

It introduces a method for classifying optics-free images using deep neural networks, eliminating the need for image reconstruction.

Findings

Achieved 92% accuracy in classification tasks

Showed potential for privacy and power efficiency improvements

Validated deep learning approach on optics-free images

Abstract

The thinnest possible camera is achieved by removing all optics, leaving only the image sensor. We train deep neural networks to perform multi-class detection and binary classification (with accuracy of 92%) on optics-free images without the need for anthropocentric image reconstructions. Inferencing from optics-free images has the potential for enhanced privacy and power efficiency.