EvFocus: Learning to Reconstruct Sharp Images from Out-of-Focus Event Streams

Event cameras are innovative sensors that capture brightness changes as asynchronous events rather than traditional intensity frames. These cameras offer substantial advantages over conventional cameras, including high temporal resolution, high dynamic range, and the elimination of motion blur. However, defocus blur, a common image quality degradation resulting from out-of-focus lenses, complicates the challenge of event-based imaging. Due to the unique imaging mechanism of event cameras, existing focusing algorithms struggle to operate efficiently on sparse event data. In this work, we propose EvFocus, a novel architecture designed to reconstruct sharp images from defocus event streams for the first time. Our work includes the development of an event-based out-of-focus camera model and a simulator to generate realistic defocus event streams for robust training and testing. EvDefous integrates a temporal information encoder, a blur-aware two-branch decoder, and a reconstruction and re-defocus module to effectively learn and correct defocus blur. Extensive experiments on both simulated and real-world datasets demonstrate that EvFocus outperforms existing methods across varying lighting conditions and blur sizes, proving its robustness and practical applicability in event-based defocus imaging.

Imagine trying to take a clear photo while running or in dim light—traditional cameras often struggle, producing blurry or dark images. Event cameras work differently: instead of recording full images at regular intervals, they detect tiny changes in light as they happen, making them ideal for fast-moving or low-light environments. They’re like having super-fast eyes that never blink. But even these advanced cameras have a blind spot: when the lens is out of focus, the data they collect becomes less useful, like trying to read a newspaper with foggy glasses.To solve this, we created EvFocus, a machine learning model that can take this fuzzy data and reconstruct sharp, clear images. It’s trained using a simulator we built to mimic how real-world blur affects event data. Our model uses both timing and blur cues—like a detective piecing together a scene from partial clues—to recover the original image.This technology could make event cameras more reliable in real-world tasks like autonomous driving, drone navigation, and security monitoring, where cameras can’t always be perfectly focused. EvFocus helps ensure these systems still ''see clearly'' when it matters most.