提出无需解码器的量化图像恢复框架，显著提升边缘设备性能
Decoder-Free Distillation for Quantized Image Restoration
- 采用无解码器蒸馏消除量化误差放大，解决低层视觉难题
- Int8模型恢复96.5%原始精度，推理速度达442帧/秒
- 适合部署在边缘设备的轻量级图像恢复任务
量化感知训练（QAT）结合知识蒸馏（KD）在模型压缩以适配边缘部署方面潜力巨大。然而，针对需要高精度恢复视觉质量的图像恢复（IR）任务，联合优化仍研究不足。直接将QAT-KD应用于低层视觉面临三大瓶颈：教师-学生模型能力不匹配、解码阶段空间误差放大，以及由量化噪声引发的重建与蒸馏损失之间的优化冲突。为此，本文提出量化感知蒸馏恢复框架（QDR），通过FP32自蒸馏消除能力差异，采用无解码器蒸馏（DFD）在网络瓶颈处精准修正量化误差，避免误差传播。同时引入可学习幅度重加权（LMR）动态平衡冲突梯度，稳定训练过程。此外设计轻量级边缘友好模型（EFM），包含可学习退化门控（LDG）以动态调节空间退化区域。在四个图像恢复任务上实验表明，所提Int8模型可恢复96.5%的FP32性能，在NVIDIA Jetson Orin上达到442帧/秒，下游目标检测任务性能提升16.3 mAP。
原文摘要 · Abstract (English)
Quantization-Aware Training (QAT), combined with Knowledge Distillation (KD), holds immense promise for compressing models for edge deployment. However, joint optimization for precision-sensitive image restoration (IR) to recover visual quality from degraded images remains largely underexplored. Directly adapting QAT-KD to low-level vision reveals three critical bottlenecks: teacher-student capacity mismatch, spatial error amplification during decoder distillation, and an optimization "tug-of-war" between reconstruction and distillation losses caused by quantization noise. To tackle these, we introduce Quantization-aware Distilled Restoration (QDR), a framework for edge-deployed IR. QDR eliminates capacity mismatch via FP32 self-distillation and prevents error amplification through Decoder-Free Distillation (DFD), which corrects quantization errors strictly at the network bottleneck. To stabilize the optimization tug-of-war, we propose a Learnable Magnitude Reweighting (LMR) that dynamically balances competing gradients. Finally, we design an Edge-Friendly Model (EFM) featuring a lightweight Learnable Degradation Gating (LDG) to dynamically modulate spatial degradation localization. Extensive experiments across four IR tasks demonstrate that our Int8 model recovers 96.5% of FP32 performance, achieves 442 frames per second (FPS) on an NVIDIA Jetson Orin, and boosts downstream object detection by 16.3 mAP
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。