ISP Distillation

TL;DR

This paper proposes a method to improve machine vision on RAW images by using knowledge distillation from models trained on processed RGB images, reducing the need for ISP processing and labeled RAW data.

Contribution

It introduces a novel knowledge distillation approach that aligns RAW image model predictions with RGB model outputs, enhancing performance without ISP overhead.

Findings

Significantly improved RAW image classification and segmentation accuracy.

Model predictions on RAW images closely match those on processed RGB images.

Reduces reliance on labeled RAW datasets and ISP computation.

Abstract

Nowadays, many of the images captured are `observed' by machines only and not by humans, e.g., in autonomous systems. High-level machine vision models, such as object recognition or semantic segmentation, assume images are transformed into some canonical image space by the camera \ans{Image Signal Processor (ISP)}. However, the camera ISP is optimized for producing visually pleasing images for human observers and not for machines. Therefore, one may spare the ISP compute time and apply vision models directly to RAW images. Yet, it has been shown that training such models directly on RAW images results in a performance drop. To mitigate this drop, we use a RAW and RGB image pairs dataset, which can be easily acquired with no human labeling. We then train a model that is applied directly to the RAW data by using knowledge distillation such that the model predictions for RAW images will be aligned with the predictions of an off-the-shelf pre-trained model for processed RGB images. Our experiments show that our performance on RAW images for object classification and semantic segmentation is significantly better than models trained on labeled RAW images. It also reasonably matches the predictions of a pre-trained model on processed RGB images, while saving the ISP compute overhead.