Efficient Robustness Assessment via Adversarial Spatial-Temporal Focus on Videos
- 29 March 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 45 (9) , 10898-10912
Abstract
Adversarial robustness assessment for video recognition models has raised concerns owing to their wide applications on safety-critical tasks. Compared with images, videos have much high dimension, which brings huge computational costs when generating adversarial videos. This is especially serious for the query-based black-box attacks where gradient estimation for the threat models is usually utilized, and high dimensions will lead to a large number of queries. To mitigate this issue, we propose to simultaneously eliminate the temporal and spatial redundancy within the video to achieve an effective and efficient gradient estimation on the reduced searching space, and thus query number could decrease. To implement this idea, we design the novel A dversarial s patial- t emporal Focus ( AstFocus ) attack on videos, which performs attacks on the simultaneously focused key frames and key regions from the inter-frames and intra-frames in the video. AstFocus attack is based on the cooperative Multi-Agent Reinforcement Learning (MARL) framework. One agent is responsible for selecting key frames, and another agent is responsible for selecting key regions. These two agents are jointly trained by the common rewards received from the black-box threat models to perform a cooperative prediction. By continuously querying, the reduced searching space composed of key frames and key regions is becoming precise, and the whole query number becomes less than that on the original video. Extensive experiments on four mainstream video recognition models and three widely used action recognition datasets demonstrate that the proposed AstFocus attack outperforms the SOTA methods, which is prevenient in fooling rate, query number, time, and perturbation magnitude at the same time.Keywords
Funding Information
- National Key R & D Program of China (2020AAA0104002)
- National Natural Science Foundation of China (62076018)
This publication has 34 references indexed in Scilit:
- ZOO2017
- Universal Adversarial Perturbations2017
- Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset2017
- Temporal Segment Networks: Towards Good Practices for Deep Action Recognition2016
- LSTM: A Search Space OdysseyIEEE Transactions on Neural Networks and Learning Systems, 2016
- Modeling Spatial-Temporal Clues in a Hybrid Deep Learning Framework for Video Classification2015
- Edge Boxes: Locating Object Proposals from Edges2014
- Partially Observable Markov Decision Processes2012
- Natural Evolution Strategies2008
- A no-reference perceptual blur metric2003