Matches in SemOpenAlex for { <https://semopenalex.org/work/W3198662695> ?p ?o ?g. }
- W3198662695 abstract "Adversarial attacks have been expanded to speaker recognition (SR). However, existing attacks are often assessed using different SR models, recognition tasks and datasets, and only few adversarial defenses borrowed from computer vision are considered. Yet,these defenses have not been thoroughly evaluated against adaptive attacks. Thus, there is still a lack of quantitative understanding about the strengths and limitations of adversarial attacks and defenses. More effective defenses are also required for securing SR systems. To bridge this gap, we present SEC4SR, the first platform enabling researchers to systematically and comprehensively evaluate adversarial attacks and defenses in SR. SEC4SR incorporates 4 white-box and 2 black-box attacks, 24 defenses including our novel feature-level transformations. It also contains techniques for mounting adaptive attacks. Using SEC4SR, we conduct thus far the largest-scale empirical study on adversarial attacks and defenses in SR, involving 23 defenses, 15 attacks and 4 attack settings. Our study provides lots of useful findings that may advance future research: such as (1) all the transformations slightly degrade accuracy on benign examples and their effectiveness vary with attacks; (2) most transformations become less effective under adaptive attacks, but some transformations become more effective; (3) few transformations combined with adversarial training yield stronger defenses over some but not all attacks, while our feature-level transformation combined with adversarial training yields the strongest defense over all the attacks. Extensive experiments demonstrate capabilities and advantages of SEC4SR which can benefit future research in SR." @default.
- W3198662695 created "2021-09-13" @default.
- W3198662695 creator A5010348592 @default.
- W3198662695 creator A5015729655 @default.
- W3198662695 creator A5023363049 @default.
- W3198662695 creator A5027425633 @default.
- W3198662695 creator A5029067537 @default.
- W3198662695 creator A5089567455 @default.
- W3198662695 date "2021-09-03" @default.
- W3198662695 modified "2023-09-24" @default.
- W3198662695 title "SEC4SR: A Security Analysis Platform for Speaker Recognition" @default.
- W3198662695 cites W111477576 @default.
- W3198662695 cites W1494198834 @default.
- W3198662695 cites W1523644733 @default.
- W3198662695 cites W1552314771 @default.
- W3198662695 cites W1575718033 @default.
- W3198662695 cites W1922655562 @default.
- W3198662695 cites W1977556410 @default.
- W3198662695 cites W2041823554 @default.
- W3198662695 cites W2046056978 @default.
- W3198662695 cites W2047555270 @default.
- W3198662695 cites W2066561607 @default.
- W3198662695 cites W2090861223 @default.
- W3198662695 cites W2109364787 @default.
- W3198662695 cites W2116013313 @default.
- W3198662695 cites W2129120544 @default.
- W3198662695 cites W2141998673 @default.
- W3198662695 cites W2145370087 @default.
- W3198662695 cites W2151965738 @default.
- W3198662695 cites W2165880886 @default.
- W3198662695 cites W2193413348 @default.
- W3198662695 cites W2225312104 @default.
- W3198662695 cites W2233657726 @default.
- W3198662695 cites W2293407848 @default.
- W3198662695 cites W2295176885 @default.
- W3198662695 cites W2407374891 @default.
- W3198662695 cites W2486441166 @default.
- W3198662695 cites W2529314575 @default.
- W3198662695 cites W2577680292 @default.
- W3198662695 cites W2612434969 @default.
- W3198662695 cites W2652009 @default.
- W3198662695 cites W2735607295 @default.
- W3198662695 cites W2747262669 @default.
- W3198662695 cites W2767951891 @default.
- W3198662695 cites W2782403400 @default.
- W3198662695 cites W2787708942 @default.
- W3198662695 cites W2872172641 @default.
- W3198662695 cites W2890964092 @default.
- W3198662695 cites W2902543210 @default.
- W3198662695 cites W2916286792 @default.
- W3198662695 cites W2938358845 @default.
- W3198662695 cites W2938445486 @default.
- W3198662695 cites W2951165433 @default.
- W3198662695 cites W2952730822 @default.
- W3198662695 cites W2962700793 @default.
- W3198662695 cites W2962717526 @default.
- W3198662695 cites W2962747881 @default.
- W3198662695 cites W2962862155 @default.
- W3198662695 cites W2962866211 @default.
- W3198662695 cites W2963001136 @default.
- W3198662695 cites W2963161670 @default.
- W3198662695 cites W2963207607 @default.
- W3198662695 cites W2963273475 @default.
- W3198662695 cites W2963557656 @default.
- W3198662695 cites W2963564844 @default.
- W3198662695 cites W2963857521 @default.
- W3198662695 cites W2963920068 @default.
- W3198662695 cites W2964121744 @default.
- W3198662695 cites W2964153729 @default.
- W3198662695 cites W2964253222 @default.
- W3198662695 cites W2964301649 @default.
- W3198662695 cites W2964311653 @default.
- W3198662695 cites W2971699529 @default.
- W3198662695 cites W2971808598 @default.
- W3198662695 cites W2972488102 @default.
- W3198662695 cites W2973252307 @default.
- W3198662695 cites W2974126847 @default.
- W3198662695 cites W2983044655 @default.
- W3198662695 cites W2984968650 @default.
- W3198662695 cites W2999401213 @default.
- W3198662695 cites W2999908512 @default.
- W3198662695 cites W3025863369 @default.
- W3198662695 cites W3026494492 @default.
- W3198662695 cites W3034214559 @default.
- W3198662695 cites W3042776162 @default.
- W3198662695 cites W3048796438 @default.
- W3198662695 cites W3084424094 @default.
- W3198662695 cites W3091896612 @default.
- W3198662695 cites W3093109544 @default.
- W3198662695 cites W3093979537 @default.
- W3198662695 cites W3103340107 @default.
- W3198662695 cites W3104032928 @default.
- W3198662695 cites W3107727058 @default.
- W3198662695 cites W3107978053 @default.
- W3198662695 cites W3109668151 @default.
- W3198662695 cites W3128911095 @default.
- W3198662695 cites W3131867595 @default.
- W3198662695 cites W3140968660 @default.
- W3198662695 cites W3153453329 @default.
- W3198662695 cites W3156272176 @default.