用环境光控制手机端图像生成，实现可审计的本地化创作。
Offline Ambient-Controlled Latent Diffusion: Architecture, Telemetry, and On-Device Evaluation
- 通过环境光传感器驱动图像生成，替代文本提示。
- 生成结果与光照强度显著相关（r=0.532），验证了可控性。
- 全链路本地运行，支持离线分析与设备端评估。
多数移动端图像生成应用依赖云端服务，导致输出难以审计。本文提出一个在 Android 设备上完全本地运行的潜在扩散模型应用，其生成过程由环境光传感器驱动，而非文本提示，确保生成、遥测和存储全程本地化。贡献不在于新扩散方法，而在于配套的测量工作流：每个生成结果均绑定传感器读数、运行路径和随机种子，形成可追溯的离线审计链。在单台三星折叠屏设备上，一次固定采集生成 373 个样本，结果显示控制器的对数光照输入与输出亮度正相关（皮尔逊相关系数 r=0.532，95% 置信区间 [0.455, 0.601]），证实光照依赖性在去噪与 VAE 解码后依然保持；潜空间 UNet/VAE 流水线在 Android Neural Networks API (NNAPI) 下，三种质量层级的平均延迟为 552–1334 毫秒。
原文摘要 · Abstract (English)
Most mobile image-generation applications are thin clients over cloud services, leaving outputs hard to audit. We present an Android latent-diffusion application that runs entirely on-device and is driven by the ambient-light sensor rather than a text prompt, keeping generation, telemetry, and storage local. The contribution is not a new diffusion method but the surrounding measurement workflow: each output is bound to the sensor reading, runtime path, and seed that produced it, giving a per-artifact audit trail for offline analysis. On a single Samsung foldable, one fixed capture of 373 artifacts shows the controller's log-lux input positively associated with output luminance (Pearson $r=0.532$, 95\% CI $[0.455, 0.601]$), confirming the ambient dependency survives denoising and VAE decoding, while the latent UNet/VAE pipeline runs at 552--1334\,ms mean latency across three quality tiers under the Android Neural Networks API (NNAPI).
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。