Reproduce “HHW-Ego: DSLR-Quality Enhancement for Multi-Source Wearable Ego Imaging” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yunfei Tong, Qianqian Sun, Jixiang Luo, Dell Zhang, Hao Sun
The authors introduce HHW-Ego, a dataset of 5,000 groups pairing wearable images with HDR references and hyperspectral data, and synthesize wearable-like training inputs by aligning color, blur, and noise. Fine-tuning Real-ESRGAN improves PSNR, SSIM, Lab distance, and LPIPS on held-out synthetic HDR pairs; hyperspectral correction improves 35 of 42 no-reference comparisons on real images, with metric-dependent evidence.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.