SIRENs are powerful implicit neural representations, but they are highly sensitive to frequency initialization: low frequencies over-smooth fine detail, while high frequencies inject noise into smooth regions. SASNet addresses this with a fixed frequency embedding paired with jointly learned spatially-adaptive masks that localize each sinusoidal neuron’s influence across the domain. The result is sharper reconstructions, less noise, and faster convergence across image fitting, 3D volumes, and SDF reconstruction — consistently outperforming INR baselines.

Jun 1, 2026

A follow-up to (CVPR 2024 INRV Workshop), advancing terrain INRs toward a compact, efficient neural terrain data format with wavelet-guided spatial adaptivity, derivative-aware supervision, and post-training compression.

May 21, 2026

Published in ACM Transactions on Spatial Algorithms and Systems (TSAS). This is the extended journal version of our SIGSPATIAL'24 conference paper , with new contributions in scale-aware seam-free sampling (S3), improved TIN smoothing, and a fully parallel GPU implementation of critical point tracking.

May 8, 2026

Accepted at CVPR 2026 Workshop PVUW. This work studies how to make VideoLLMs more camera-aware by benchmarking camera motion understanding and injecting geometry-derived motion cues at inference time.

Mar 13, 2026