AI
FurE: Efficient Instance-Specific 3D Fur Reconstruction without Animal-Fur Datasets
Publication: Hugging Face Daily Papers
Author: Not available in collected metadata
Source published: 27/09/2026, 20:00:00 UTC
Briefing created: 30/09/2026, 04:43:26 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 30/09/2026, 02:18:24 UTC.
Briefings
FurE is a new method for reconstructing animal fur from multi-view images, achieving 10x speedup over existing SOTA dense per-strand optimization. It uses a root-conditioned latent field optimized via a PCA-based decoder, which can be edited by users. The method leverages local fur-thickness cues and part-based priors to reconstruct a defurred body, demonstrating that human-hair strand data can substitute for scarce animal-fur datasets. Validation shows the method works on both synthetic and real-world sequences while maintaining strand fidelity.
Key points
- FurE achieves a 10x speedup in strand training compared to current state-of-the-art dense per-strand optimization methods.
- The method uses a PCA-based decoder trained on human-hair strand data to replace the need for animal-fur datasets.
Context and limitations
Validation is limited to synthetic and real-world sequences; quantitative and qualitative results are presented despite the reduction in training time.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 30/09/2026, 04:43:26 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeiejmjmbt6fwr75vjh4ejviksl4rv6va32sb7pwxrivykima6jhjiq