BADiff: Bandwidth Adaptive Diffusion Model

TL;DR

BADiff introduces a bandwidth-aware diffusion model that adaptively modulates image generation quality based on real-time network constraints, improving visual fidelity in bandwidth-limited scenarios.

Contribution

It presents a novel training strategy for diffusion models to dynamically adjust denoising based on bandwidth, with minimal architectural changes.

Findings

Significantly improves visual quality under bandwidth constraints.

Enables early-stopping sampling guided by a lightweight quality embedding.

Achieves better results than naive early-stopping methods.

Abstract

In this work, we propose a novel framework to enable diffusion models to adapt their generation quality based on real-time network bandwidth constraints. Traditional diffusion models produce high-fidelity images by performing a fixed number of denoising steps, regardless of downstream transmission limitations. However, in practical cloud-to-device scenarios, limited bandwidth often necessitates heavy compression, leading to loss of fine textures and wasted computation. To address this, we introduce a joint end-to-end training strategy where the diffusion model is conditioned on a target quality level derived from the available bandwidth. During training, the model learns to adaptively modulate the denoising process, enabling early-stop sampling that maintains perceptual quality appropriate to the target transmission condition. Our method requires minimal architectural changes and leverages a lightweight quality embedding to guide the denoising trajectory. Experimental results demonstrate that our approach significantly improves the visual fidelity of bandwidth-adapted generations compared to naive early-stopping, offering a promising solution for efficient image delivery in bandwidth-constrained environments. Code is available at: https://github.com/xzhang9308/BADiff.