Affiliation:
1. University of California Los Angeles, California, USA
2. University of Southern California, Information Sciences Institute, California, USA
Abstract
End-to-end deep learning models are increasingly applied to safety-critical human activity recognition (HAR) applications, e.g., healthcare monitoring and smart home control, to reduce developer burden and increase the performance and robustness of prediction models. However, integrating HAR models in safety-critical applications requires trust, and recent approaches have aimed to balance the performance of deep learning models with explainable decision-making for complex activity recognition. Prior works have exploited the compositionality of complex HAR (i.e., higher-level activities composed of lower-level activities) to form models with symbolic interfaces, such as concept-bottleneck architectures, that facilitate inherently interpretable models. However, feature engineering for symbolic concepts-as well as the relationship between the concepts-requires precise annotation of lower-level activities by domain experts, usually with fixed time windows, all of which induce a heavy and error-prone workload on the domain expert. In this paper, we introduce X-CHAR, an eXplainable Complex Human Activity Recognition model that doesn't require precise annotation of low-level activities, offers explanations in the form of human-understandable, high-level concepts, while maintaining the robust performance of end-to-end deep learning models for time series data. X-CHAR learns to model complex activity recognition in the form of a sequence of concepts. For each classification, X-CHAR outputs a sequence of concepts and a counterfactual example as the explanation. We show that the sequence information of the concepts can be modeled using Connectionist Temporal Classification (CTC) loss without having accurate start and end times of low-level annotations in the training dataset-significantly reducing developer burden. We evaluate our model on several complex activity datasets and demonstrate that our model offers explanations without compromising the prediction accuracy in comparison to baseline models. Finally, we conducted a mechanical Turk study to show that the explanations provided by our model are more understandable than the explanations from existing methods for complex activity recognition.
Funder
the NIH mHealth Center for Discovery, Optimization and Translation of Temporally-Precise Interventions
Army Research Laboratory
CONIX Research Center
AFOSR
NSF
Publisher
Association for Computing Machinery (ACM)
Subject
Computer Networks and Communications,Hardware and Architecture,Human-Computer Interaction
Reference65 articles.
1. 2016. Defense Advanced Research Projects Agency . Broad Agency Announcement , Explainable Artificial Intelligence (XAI). https://www.darpa.mil/attachments/DARPA-BAA-16-53.pdf. Online; accessed 14-November-2022. 2016. Defense Advanced Research Projects Agency. Broad Agency Announcement, Explainable Artificial Intelligence (XAI). https://www.darpa.mil/attachments/DARPA-BAA-16-53.pdf. Online; accessed 14-November-2022.
2. 2018. Article 15 EU GDPR "Right of access by the data subject". https://www.privacy-regulation.eu/en/article-15-right-of-access-by-the-data-subject-GDPR.htm. Online ; accessed 04- March - 2022 . 2018. Article 15 EU GDPR "Right of access by the data subject". https://www.privacy-regulation.eu/en/article-15-right-of-access-by-the-data-subject-GDPR.htm. Online; accessed 04-March-2022.
3. 2018. Recital 71 EU GDPR. https://www.privacy-regulation.eu/en/r71.htm. Online ; accessed 04- March - 2022 . 2018. Recital 71 EU GDPR. https://www.privacy-regulation.eu/en/r71.htm. Online; accessed 04-March-2022.
4. mORAL
5. DeXAR
Cited by
13 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献