Nexus: Structured Synergy for Efficient Text-to-Image Generation using Rectified Flow Model
2026-08-17 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed Nexus, a new way to create images from text that uses less computer power and memory than current methods. By combining special layers called MoE feed-forward layers, gated DeltaNet attention, and training that uses fewer bits, Nexus works faster and uses less memory. Their tests on standard image datasets show that Nexus produces image quality similar to popular models but runs much more efficiently. This could help with using image generation on devices with limited resources.
diffusion modelsflow matchingMoE (Mixture of Experts)DeltaNet attentionlow-bit quantizationlinear complexitytext-to-image generationinference efficiencyCOCO datasetLAION dataset
Authors
Yizhao Wang
Abstract
Diffusion and flow matching models have made significant progress in text-to-image generation, yet high computation, quadratic complexity, and large memory footprint hinder high-resolution synthesis and edge deployment. We propose Nexus, which integrates sparse architecture, linear complexity, and low-bit quantization. It combines MoE feed-forward layers, gated DeltaNet attention, and per-expert low-bit training to reduce computation and memory. Their joint optimization allows Nexus to achieve generation quality comparable to mainstream models such as SDXL and SD3 while delivering markedly higher inference efficiency. Experiments on COCO and LAION validate its effectiveness.