Learning Digital Camera Pipeline for Extreme Low-Light Imaging
This addresses the issue of poor image quality for users in low-light environments, representing a strong specific gain rather than a broad paradigm shift.
The paper tackles the problem of producing dark and noisy images in extreme low-light conditions by learning a camera pipeline to transform RAW sensor data into well-exposed sRGB images, resulting in significant improvements in visual quality as shown by outperforming state-of-the-art methods in psychophysical tests and various metrics.
In low-light conditions, a conventional camera imaging pipeline produces sub-optimal images that are usually dark and noisy due to a low photon count and low signal-to-noise ratio (SNR). We present a data-driven approach that learns the desired properties of well-exposed images and reflects them in images that are captured in extremely low ambient light environments, thereby significantly improving the visual quality of these low-light images. We propose a new loss function that exploits the characteristics of both pixel-wise and perceptual metrics, enabling our deep neural network to learn the camera processing pipeline to transform the short-exposure, low-light RAW sensor data to well-exposed sRGB images. The results show that our method outperforms the state-of-the-art according to psychophysical tests as well as pixel-wise standard metrics and recent learning-based perceptual image quality measures.