面向轻量级深度伪造检测的无数据模型压缩  

Data-free model compression for light-weight DeepFake detection

作　　者：卓文琦 李东泽 王伟[2] 董晶[2] Zhuo Wenqi;Li Dongze;Wang Wei;Dong Jing(School of Artificial Intelligence,University of Chinese Academy of Sciences,Beijing 100049,China;Center for Research on Intelligent Perception and Computing,Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China)

机构地区：[1]中国科学院大学人工智能学院,北京100049 [2]中国科学院自动化研究所智能感知与计算研究中心,北京100190

出　　处：《中国图象图形学报》2023年第3期820-835,共16页Journal of Image and Graphics

基　　金：国家重点研发计划资助(2020AAA0140003);国家自然科学基金项目(61972395)。

摘　　要：目的尽管现有的深度伪造检测方法已在各大公开数据集上展现出了极佳的真伪鉴别性能,但考虑到运行过程中耗费的巨大内存占用和计算成本,如何实现此类模型的在线部署仍是一个具有挑战性的任务。对此,本文尝试利用无数据量化的方法开发轻量级的深度伪造检测器。方法在保证准确率损失较少的前提下,对提前训练好的高精度深度伪造检测模型进行压缩处理,不再使用32 bit浮点数表示模型的权重参数与激活值,而是将其全部转化为低位宽的整型数值。此外,由于人脸数据涉及隐私保护问题,本文中所有的量化操作都是基于无数据场景完成的,即利用合成数据作为校准集来获取正确的激活值范围。这些数据经过不断优化迭代,完美匹配了存储在预训练模型各批归一化层中的统计信息,与原始训练数据具备非常相似的分布特征。结果在两个经典的人脸伪造数据集Face Forensics++和Celeb-DF v2上,4种预先训练好的深度伪造检测模型Res Net50、Xception、EfficientNet-b3和MobileNetV2经过所提方法的量化压缩处理后,均能保持甚至超越原有的性能指标。即使当模型的权重和激活值被压缩为6 bit时,所得轻量级模型的最低检测准确率也能达到81%。结论通过充分利用蕴含在深度伪造检测预训练模型中的有价值信息,本文提出了一种基于无数据模型压缩的轻量级人脸伪造检测器,该检测器能够准确高效地识别出可疑人脸样本的真实性,与此同时,检测所需的资源和时间成本大幅降低。Objective Deep generative models-based human facial images and videos analyses have been developing in recent years.To cope with the faked issues effectively,a novel DeepFake detection(DFD)technique has emerged.Multiple DFD methods are yielded the detector to discriminate between the real and fake faces analysis with over 95%precision.However,it is still a great challenge to deploy them online because of the memory and computational cost.So,we develop a quantified model to DFD domain.Quantization-related model compression can be used to optimize model size through converting a model’s key parameters from high precision floating points into low precision integers.However,the degradation issue is still being challenged.To resolve degradation problem,it can be segmented into 2 categories:1)quantification-oriented fine-tuning and 2)post-training quantification.To optimize cost effective,the latter one is optioned to develop a light-weight DFD detector.In addition,to clarify the privacy concerns and information security,data-free scenario-oriented models-quantified are constructed and optimized with no prior training set.Method The proposed framework consists of 2 steps:1)key parameters-related quantification and 2)activation-ranged calibration.First,the weights and activations of a well-trained high accuracy DFD model are optioned as the target parameters to be quantified.A linear transformation-asymmetric is used to convert them from 32-bit floating points into lower bit-width representation like INT8 and INT6.Next,the activation-ranged errors are validated based on calibration set.For data-free scenario,it is challenged to collect data from prior training set.Therefore,to produce more effective calibration data,a batch of normalization layers of a pre-trained DFD model is tailored to guide the generator.Such statistics knowledge is often used to reflect the distribution of training data like running-relevant means and variances.We can optimize the input data of those are sampled in random from a standard Gaussian

关 键 词：深度伪造检测 虚假人脸 模型压缩 低位宽表示 无数据蒸馏 轻量级模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]