xiao zhang
@xiaozha55937919
UChicago CS PhD Computer Vision
Excited to be at @CVPR 2025! Looking forward to catching up with old friends and meeting new ones. If you're interested in grabbing coffee, trying out new restaurants, or chatting about generative representation learning, feel free to DM me! A quick summary of my recent works:…
In case you missed it, our recent work shows that when you train a model for relighting, you get albedos for free from its latent features without needing to see albedo-like images. Not a single labeled image is needed. Completely zero-shot! It emerges as a byproduct of this…
[1/6] Super excited to share our latest work, "Latent Intrinsics Emerge When Training to Relight." We found a surprisingly easy method to extract albedo without seeing albedo-like images. We questioned why intrinsic image representations must be three-channel maps. We've been…
[1/6] Super excited to share our latest work, "Latent Intrinsics Emerge When Training to Relight." We found a surprisingly easy method to extract albedo without seeing albedo-like images. We questioned why intrinsic image representations must be three-channel maps. We've been…