Track trials, findings, and clinical evidence as it lands.
Search papers, labs, and topics across Lattice.
ExpandDiff addresses single-image HDR reconstruction by employing a conditional diffusion pipeline to jointly recover clipped shadows and highlights, a challenge arising from varying sensor dynamic ranges and exposures. The method utilizes Dynamic Clipping Synthesis (DCS) to generate diverse training data by randomly sampling clipping percentiles, and a pixel-space diffusion model with spatially-adaptive normalization and a bounded output head to predict perceptually encoded HDR. This approach achieves state-of-the-art results, significantly outperforming prior methods on the SI-HDR benchmark, particularly under two-sided clipping conditions.
A single diffusion model can simultaneously reconstruct clipped shadows and highlights in LDR images, achieving over 7 dB improvement in PU21-PSNR under two-sided clipping.
Single-image HDR reconstruction requires inferring missing detail while preserving the visible content of an LDR image. Differences in sensor dynamic range and exposure cause LDR images to lose varying amounts of information in shadows and highlights. We present ExpandDiff, a conditional diffusion pipeline that jointly reconstructs clipped shadows and highlights. To account for this variation, we introduce Dynamic Clipping Synthesis (DCS), which randomly samples shadow and highlight clipping percentiles when constructing training inputs from HDR targets. A pixel-space diffusion model guided by spatially-adaptive normalization then predicts perceptually encoded HDR through a bounded output head, reconstructing both clipping directions in one sampling trajectory. On the SI-HDR benchmark, ExpandDiff variants improve HDR reconstruction accuracy by 3.43 dB in PU21-PSNR over the strongest evaluated competing method, and by 7.34 dB under two-sided clipping. The code and supplementary material are available at https://memreandiran.github.io/expanddiff/.