Transferable adversarial examples cause practical security risks since they can mislead a target model without knowing its internal knowledge. A conventional recipe for maximizing transferability is to keep only the optimal adversarial example from all those obtained in the optimization pipeline. In this paper, for the first time, we revisit this convention and demonstrate that those discarded, sub-optimal adversarial examples can be reused to boost transferability. Specifically, we propose “Adversarial Example Soups” (AES), with AES-tune for averaging discarded adversarial examples in hyperparameter tuning and AES-rand for stability testing. In addition, our AES is inspired by “model soups”, which averages weights of multiple fine-tuned models for improved accuracy without increasing inference time. Extensive experiments validate the global effectiveness of our AES, boosting 10 state-of-the-art transfer attacks and their combinations by up to 13% against 10 diverse (defensive) target models. We also show the possibility of generalizing AES to other types, e.g., directly averaging multiple in-the-wild adversarial examples that yield comparable success. A promising byproduct of AES is the improved stealthiness of adversarial examples since the perturbation variances are naturally reduced.
Automatic Speech Recognition (ASR) provides a new way of human-computer interaction. However, it is vulnerable to adversarial examples, which are obtained by deliberately adding perturbations to the original audios. Thorough studies on the universal feature of adversarial examples are essential to prevent potential attacks. Previous research has shown classic adversarial examples have different logits distribution compared to normal speech. This paper proposes a logit-traction attack to eliminate this difference at the statistical level. Experiments on the LibriSpeech dataset show that the proposed attack reduces the accuracy of the LOGITS NOISE detection to 52.1%. To further verify the effectiveness of this approach in attacking detection based on logits, three different features quantifying the dispersion of logits are constructed in this paper. Furthermore, a richer target sentence is adopted for experiments. The results indicate that these features can detect baseline adversarial examples with an accuracy of about 90% but cannot effectively detect Logits-Traction adversarial examples, proving that Logits-Traction attack can evade the logits-based detection method.
Through experiments on combinations of coding tools included in H.264/AVC, we obtain an optimal combination of coding tools for the mobile multimedia broadcasting application. Although H.264/AVC has defined several profiles and levels for different applications, the partitions are still imprecise. The mobile communication network is usually with limited bandwidth and different kinds of fading and multipath interference. The mobile terminals with downlink video service in the broadcasting network are battery restricted and low processing power. Considering of these features, the proposal has made an optimal selection of the coding tools in H.264/AVC for the mobile multimedia broadcasting application. This set of tools makes a good balance among compression efficiency, transmission robustness and decoding complexity, therefore provides reliable quality for mobile multimedia broadcasting service.