Measurement-Adapted Eigentask Representations for Photon-Limited Optical Readout
2026-05-11 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionEmerging Technologies
AI summaryⓘ
The authors studied how to better represent noisy optical sensor data to improve decisions made from low-light images. They introduced "eigentasks," a way to organize measurement features by how clearly they can be seen despite noise. By testing real optical data and existing neural network data, they found eigentasks often performed better than common methods like PCA. This improvement was most notable when there are few light photons, limited training data, or harder classification tasks. Their approach helps extract useful information more efficiently from noisy optical measurements.
optical readoutphoton shot noisedetector noisequantization erroreigentasksprincipal component analysislow-light imagingfew-shot learningclassificationmeasurement noise
Authors
Tianyang Chen, Mandar M. Sohoni, Saeed A. Khan, Jérémie Laydevant, Shi-Yuan Ma, Tianyu Wang, Peter L. McMahon, Hakan E. Türeci
Abstract
Optical readout in low-light imaging is fundamentally limited by measurement noise, including photon shot noise, detector noise, and quantization error. In this regime, downstream inference depends not only on the optical front end, but also on how noisy high-dimensional sensor measurements are represented before classification or decision-making. Here we show that eigentasks provide a measurement-adapted representation for optical sensor outputs by ordering readout features according to their resolvability under noise. Using experimental data from a lens-based optical imaging system and a reanalysis of published data from a single-photon-detection neural network, we find that eigentask representations frequently outperform standard baselines including principal component analysis and filtering-based compression. The advantage is most pronounced in photon-limited, few-shot, and higher-difficulty classification regimes. In few-shot MPEG-7 classification, for example, the advantage over other methods reaches about 10 percentage points as the number of classes increases. In these settings, eigentasks yield more informative low-dimensional features and improve sample-efficient downstream learning. These results identify measurement-adapted representation as a promising strategy for optical inference when photon budget, acquisition time, and task complexity are constrained.