WaiT: Wavelet-Aware Flow-Matching Sets New SOTA, Halves Compute
CSProfKGD · x · 2026-08-04
The paper WaiT for the Signal introduces a Wavelet-aware image Transformer (WaiT). While standard flow matching treats all spatial frequencies uniformly, WaiT decomposes generation into coarse and fine bands via lossless wavelets.
Its core mechanism makes high-frequency bands 'wait' for the signal: staying as pure noise until coarse structures emerge, then joining the flow for joint refinement. Experiments show that on ImageNet 512x512, this method achieves a pixel-space FID of 1.43, is Pareto-optimal, and reduces sampling compute by up to 50%. With a 2B parameter model, it sets a new SOTA FID of 1.3. The architecture also scales seamlessly to high-res OpenImages and video generation, achieving an SOTA FVD of 0.84 on Kinetics-600.
More from Multimodal
- Google Earth Pauses AI Feature After Disinformation Nightmare — nordicinst · 2026-08-04
- Elon Musk Showcases Grok Imagine's Architectural Design Capabilities — elonmusk · 2026-08-04
- Speeding Up MiniMax H3: Useful ComfyUI Node Combinations — fruesome · 2026-08-04
- MiniMax H3 Time-Saver: 0.2MP Resolution Draft Mode — Independent-Frequent · 2026-08-04
- RTX 3080 Benchmark: Spectrum Nodes Significantly Accelerate MiniMax H3 Video Generation — Valuable_Issue_ · 2026-08-04
- Creating an Indian Metal Band Music Video Using Midjourney and Seedance — No-Accountant-3705 · 2026-08-04