{"project":[{"call_identifier":"FWF","grant_number":"Z211","name":"The Wittgenstein Prize","_id":"25F42A32-B435-11E9-9278-68D0E5697425"}],"isi":1,"acknowledgement":"We thank Robert Geirhos and Roland Zimmermann for their participation in the case study and valuable feedback, Chris Olah and Nick Cammarata for valuable discussions in the early phase of the project, as well as the Distill Slack workspace as a platform for discussions. M.L. is supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award). J.B. is supported by the German Federal Ministry of Education and Research\r\n(BMBF) through the Competence Center for Machine Learning (TUE.AI, FKZ 01IS18039A) and the International Max Planck Research School for Intelligent Systems (IMPRS-IS). R.H. is partially supported by Boeing and Horizon-2020 ECSEL (grant 783163, iDev40).\r\n","year":"2021","volume":40,"intvolume":"        40","language":[{"iso":"eng"}],"_id":"10404","article_type":"original","citation":{"ieee":"S. Sietzen, M. Lechner, J. Borowski, R. Hasani, and M. Waldner, “Interactive analysis of CNN robustness,” <i>Computer Graphics Forum</i>, vol. 40, no. 7. Wiley, pp. 253–264, 2021.","mla":"Sietzen, Stefan, et al. “Interactive Analysis of CNN Robustness.” <i>Computer Graphics Forum</i>, vol. 40, no. 7, Wiley, 2021, pp. 253–64, doi:<a href=\"https://doi.org/10.1111/cgf.14418\">10.1111/cgf.14418</a>.","short":"S. Sietzen, M. Lechner, J. Borowski, R. Hasani, M. Waldner, Computer Graphics Forum 40 (2021) 253–264.","ama":"Sietzen S, Lechner M, Borowski J, Hasani R, Waldner M. Interactive analysis of CNN robustness. <i>Computer Graphics Forum</i>. 2021;40(7):253-264. doi:<a href=\"https://doi.org/10.1111/cgf.14418\">10.1111/cgf.14418</a>","ista":"Sietzen S, Lechner M, Borowski J, Hasani R, Waldner M. 2021. Interactive analysis of CNN robustness. Computer Graphics Forum. 40(7), 253–264.","apa":"Sietzen, S., Lechner, M., Borowski, J., Hasani, R., &#38; Waldner, M. (2021). Interactive analysis of CNN robustness. <i>Computer Graphics Forum</i>. Wiley. <a href=\"https://doi.org/10.1111/cgf.14418\">https://doi.org/10.1111/cgf.14418</a>","chicago":"Sietzen, Stefan, Mathias Lechner, Judy Borowski, Ramin Hasani, and Manuela Waldner. “Interactive Analysis of CNN Robustness.” <i>Computer Graphics Forum</i>. Wiley, 2021. <a href=\"https://doi.org/10.1111/cgf.14418\">https://doi.org/10.1111/cgf.14418</a>."},"publication_identifier":{"issn":["0167-7055"],"eissn":["1467-8659"]},"external_id":{"isi":["000722952000024"],"arxiv":["2110.07667"]},"doi":"10.1111/cgf.14418","page":"253-264","publication":"Computer Graphics Forum","author":[{"last_name":"Sietzen","first_name":"Stefan","full_name":"Sietzen, Stefan"},{"first_name":"Mathias","last_name":"Lechner","id":"3DC22916-F248-11E8-B48F-1D18A9856A87","full_name":"Lechner, Mathias"},{"first_name":"Judy","last_name":"Borowski","full_name":"Borowski, Judy"},{"last_name":"Hasani","first_name":"Ramin","full_name":"Hasani, Ramin"},{"first_name":"Manuela","last_name":"Waldner","full_name":"Waldner, Manuela"}],"status":"public","title":"Interactive analysis of CNN robustness","quality_controlled":"1","scopus_import":"1","publisher":"Wiley","user_id":"4359f0d1-fa6c-11eb-b949-802e58b17ae8","date_created":"2021-12-05T23:01:40Z","oa":1,"oa_version":"Preprint","abstract":[{"text":"While convolutional neural networks (CNNs) have found wide adoption as state-of-the-art models for image-related tasks, their predictions are often highly sensitive to small input perturbations, which the human vision is robust against. This paper presents Perturber, a web-based application that allows users to instantaneously explore how CNN activations and predictions evolve when a 3D input scene is interactively perturbed. Perturber offers a large variety of scene modifications, such as camera controls, lighting and shading effects, background modifications, object morphing, as well as adversarial attacks, to facilitate the discovery of potential vulnerabilities. Fine-tuned model versions can be directly compared for qualitative evaluation of their robustness. Case studies with machine learning experts have shown that Perturber helps users to quickly generate hypotheses about model vulnerabilities and to qualitatively compare model behavior. Using quantitative analyses, we could replicate users’ insights with other CNN architectures and input images, yielding new insights about the vulnerability of adversarially trained models.","lang":"eng"}],"main_file_link":[{"url":"https://arxiv.org/abs/2110.07667","open_access":"1"}],"department":[{"_id":"ToHe"}],"issue":"7","date_published":"2021-11-27T00:00:00Z","publication_status":"published","date_updated":"2023-08-14T13:11:42Z","article_processing_charge":"No","month":"11","day":"27","type":"journal_article"}