Search papers, labs, and topics across Lattice.
Department of Computer Science, City University of Hong Kong
3
0
7
CanonicalGS achieves a remarkable 2.5 dB improvement in novel view synthesis quality by stabilizing scene representations against noisy input data.
A new dataset, SeIQA, offers a benchmark to evaluate how humans perceive semantic loss in degraded images, pushing beyond traditional quality metrics.
Domain-specific speech recognition can be both accurate AND fast: EndoASR slashes error rates for endoscopy reports while running 10x faster than Whisper-large-v3.
