Undergrad student, The Australian National University
4 papers at NeurIPS 2025
ZPressor is an architecture-agnostic module that compresses multi-view inputs for scalable feed-forward 3DGS.
FlashMo introduces a geometric factorized interpolant and frequency-sparse attention, enabling scalable efficient 3D motion diffusion.
A novel semi-supervised learning paradigm that unifies view-wise co-training, meta-learned supervision, and adversarial perturbation through a structured triadic game.
FPSAttention is a training-aware FP8 quantization and sparsity co-design for video diffusion models that achieves up to 4.96× speedup without quality loss by aligning 3D tile granularity, denoising-step adaptation, and hardware-efficient kernels.
