More Like this

1. A Target-Agnostic Attack on Deep Models: Exploiting Security Vulnerabilities of Transfer Learning

   Rezaei, Shahbaz; Liu, Xin (January 2020, The International Conference on Learning Representations)

   Due to insufficient training data and the high computational cost to train a deep neural network from scratch, transfer learning has been extensively used in many deep-neural-network-based applications. A commonly used transfer learning approach involves taking a part of a pre-trained model, adding a few layers at the end, and re-training the new layers with a small dataset. This approach, while efficient and widely used, imposes a security vulnerability because the pre-trained model used in transfer learning is usually publicly available, including to potential attackers. In this paper, we show that without any additional knowledge other than the pre-trained model, an attacker can launch an effective and efficient brute force attack that can craft instances of input to trigger each target class with high confidence. We assume that the attacker has no access to any target-specific information, including samples from target classes, re-trained model, and probabilities assigned by Softmax to each class, and thus making the attack target-agnostic. These assumptions render all previous attack models inapplicable, to the best of our knowledge. To evaluate the proposed attack, we perform a set of experiments on face recognition and speech recognition tasks and show the effectiveness of the attack. Our work reveals a fundamental security weakness of the Softmax layer when used in transfer learning settings. 

   more » « less
2. On the Learning Property of Logistic and Softmax Losses for Deep Neural Networks

   https://doi.org/10.1609/aaai.v34i04.5907  

   Li, Xiangrui; Li, Xin; Pan, Deng; Zhu, Dongxiao (June 2020, Proceedings of the AAAI Conference on Artificial Intelligence)

   Deep convolutional neural networks (CNNs) trained with logistic and softmax losses have made significant advancement in visual recognition tasks in computer vision. When training data exhibit class imbalances, the class-wise reweighted version of logistic and softmax losses are often used to boost performance of the unweighted version. In this paper, motivated to explain the reweighting mechanism, we explicate the learning property of those two loss functions by analyzing the necessary condition (e.g., gradient equals to zero) after training CNNs to converge to a local minimum. The analysis immediately provides us explanations for understanding (1) quantitative effects of the class-wise reweighting mechanism: deterministic effectiveness for binary classification using logistic loss yet indeterministic for multi-class classification using softmax loss; (2) disadvantage of logistic loss for single-label multi-class classification via one-vs.-all approach, which is due to the averaging effect on predicted probabilities for the negative class (e.g., non-target classes) in the learning process. With the disadvantage and advantage of logistic loss disentangled, we thereafter propose a novel reweighted logistic loss for multi-class classification. Our simple yet effective formulation improves ordinary logistic loss by focusing on learning hard non-target classes (target vs. non-target class in one-vs.-all) and turned out to be competitive with softmax loss. We evaluate our method on several benchmark datasets to demonstrate its effectiveness. 

   more » « less
3. AdaFilter: Adaptive Filter Fine-Tuning for Deep Transfer Learning

   https://doi.org/10.1609/aaai.v34i04.5824  

   Guo, Yunhui; Li, Yandong; Wang, Liqiang; Rosing, Tajana (June 2020, Proceedings of the AAAI Conference on Artificial Intelligence)

   There is an increasing number of pre-trained deep neural network models. However, it is still unclear how to effectively use these models for a new task. Transfer learning, which aims to transfer knowledge from source tasks to a target task, is an effective solution to this problem. Fine-tuning is a popular transfer learning technique for deep neural networks where a few rounds of training are applied to the parameters of a pre-trained model to adapt them to a new task. Despite its popularity, in this paper we show that fine-tuning suffers from several drawbacks. We propose an adaptive fine-tuning approach, called AdaFilter, which selects only a part of the convolutional filters in the pre-trained model to optimize on a per-example basis. We use a recurrent gated network to selectively fine-tune convolutional filters based on the activations of the previous layer. We experiment with 7 public image classification datasets and the results show that AdaFilter can reduce the average classification error of the standard fine-tuning by 2.54%. 

   more » « less
4. Cross-Domain Unseen Activity Recognition Using Transfer Learning

   https://doi.org/10.1109/COMPSAC54236.2022.00117  

   Al Hafiz Khan, Md Abdullah; Roy, Nirmalya (June 2022, 2022 IEEE 46th Annual Computers, Software, and Applications Conference (COMPSAC))

   Activity Recognition (AR) models perform well with a large number of available training instances. However, in the presence of sensor heterogeneity, sensing biasness and variability of human behaviors and activities and unseen activity classes pose key challenges to adopting and scaling these pre-trained activity recognition models in the new environment. These challenging unseen activities recognition problems are addressed by applying transfer learning techniques that leverage a limited number of annotated samples and utilize the inherent structural patterns among activities within and across the source and target domains. This work proposes a novel AR framework that uses the pre-trained deep autoencoder model and generates features from source and target activity samples. Furthermore, this AR frame-work establishes correlations among activities between the source and target domain by exploiting intra- and inter-class knowledge transfer to mitigate the number of labeled samples and recognize unseen activities in the target domain. We validated the efficacy and effectiveness of our AR framework with three real-world data traces (Daily and Sports, Opportunistic, and Wisdm) that contain 41 users and 26 activities in total. Our AR framework achieves performance gains ≈ 5-6% with 111, 18, and 70 activity samples (20 % annotated samples) for Das, Opp, and Wisdm datasets. In addition, our proposed AR framework requires 56, 8, and 35 fewer activity samples (10% fewer annotated examples) for Das, Opp, and Wisdm, respectively, compared to the state-of-the-art Untran model. 

   more » « less
5. Scaling Human Activity Recognition via Deep Learning-based Domain Adaptation

   https://doi.org/10.1109/PERCOM.2018.8444585  

   Khan, Md Abdullah; Roy, Nirmalya; Misra, Archan (March 2018, 2018 IEEE International Conference on Pervasive Computing and Communications (PerCom))

   We investigate the problem of making human activity recognition (AR) scalable-i.e., allowing AR classifiers trained in one context to be readily adapted to a different contextual domain. This is important because AR technologies can achieve high accuracy if the classifiers are trained for a specific individual or device, but show significant degradation when the same classifier is applied context-e.g., to a different device located at a different on-body position. To allow such adaptation without requiring the onerous step of collecting large volumes of labeled training data in the target domain, we proposed a transductive transfer learning model that is specifically tuned to the properties of convolutional neural networks (CNNs). Our model, called HDCNN, assumes that the relative distribution of weights in the different CNN layers will remain invariant, as long as the set of activities being monitored does not change. Evaluation on real-world data shows that HDCNN is able to achieve high accuracy even without any labeled training data in the target domain, and offers even higher accuracy (significantly outperforming competitive shallow and deep classifiers) when even a modest amount of labeled training data is available. 

   more » « less