EdgeDiT: Hardware-Aware Diffusion Transformers for Efficient On-Device Image Generation

1. 1Design hardware-efficient architectures for diffusion transformers
2. 2Optimize computational complexity and memory requirements for edge devices
3. 3Implement EdgeDiT on mobile Neural Processing Units (NPUs) such as Qualcomm Hexagon and Apple Neural Engine

AI engineers and researchers working on computer vision and edge devices can benefit from EdgeDiT, as it enables efficient image generation on resource-constrained devices

💡 Hardware-aware design can significantly improve the efficiency of diffusion transformers on edge devices

💡 EdgeDiT enables efficient on-device image generation with hardware-aware diffusion transformers