A new framework called Sparse Surface Understanding Framework (SSUF) has been developed to improve material recognition from incomplete visual data. SSUF adapts four pre-trained architectures—ConvAE, ViT, Swin Transformer, and MAE—for simultaneous surface reconstruction and material classification. Experiments on the Touch-and-Go dataset, using only 10% of visible image data, showed that Swin Transformer achieved the highest classification accuracy at 89.21%, while MAE excelled in reconstruction. ViT offered a balanced performance, and all models demonstrated real-time inference capabilities. AI
IMPACT This research could improve robotic perception and environmental understanding by enabling material recognition from severely limited visual input.
RANK_REASON The cluster contains a research paper detailing a new framework and experimental results for a computer vision task. [lever_c_demoted from research: ic=1 ai=1.0]
- arXiv
- ConvAE
- Convolutional autoencoder and conditional random fields hybrid for predicting spatial-temporal chaos
- Mae
- Masked Autoencoder
- Sparse Surface Understanding Framework
- Swin Transformer
- Touch-and-Go dataset
- vision transformer
- Vít
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
