# A Novel Visual Interpretability for Deep Neural Networks by Optimizing Activation Maps with Perturbation

> Research article (Proceedings of the AAAI Conference on Artificial Intelligence, 2021) · cited 25× · AI/ML

**Wikidata**: [openalex:W3174863366](https://www.wikidata.org/wiki/openalex:W3174863366)  
**Source**: https://4ort.xyz/entity/a-novel-visual-interpretability-for-deep-neural-networks-by-optimizing-activation-maps-with-perturbation