Open Access
Review
Table 1
Different surveys on adversarial examples
Survey | Year | Venue | Attack? | Defense? | SOTA | Focus | Perspective | Highlights |
---|---|---|---|---|---|---|---|---|
Serban et al[14] | 2020 | CSUR | √ | √ | 2020 | Object Detection |
Neutral | Aware of earlier adversarial examples in machine learning |
Machado et al[15] | 2021 | CSUR | √ | √ | 2021 | General CV |
Defender | Instructive for defense design |
Long et al[16] | 2022 | C&S | √ | × | 2022 | General CV |
Attacker | Abundant visualization results |
Wang et al[17] | 2022 | Neural Comput. | √ | √ | 2022 | General CV |
Neutral | Comprehensive taxonomy of attacks and defenses |
Li et al[18] | 2024 | CSUR | √ | × | 2023 | 2D&3D CV |
Attacker | Considering 3D vision tasks |
Costa et al[19] | 2024 | ACCESS | √ | √ | 2024 | General CV |
Neutral | Including SOTA performance; Considering transformers |
Ours | 2024 | - | √ | √ | 2024 | General CV |
Neutral | Discussing critical issues beyond attack and defense |
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.