ExplAInable Pixels: Investigating One-Pixel Attacks on Deep Learning Models with Explainable Visualizations.
Jonas KeppelJonathan LiebersJonas AudaUwe GruenefeldStefan SchneegassPublished in: MUM (2022)
Keyphrases
- learning models
- image pixels
- neighboring pixels
- machine learning
- input image
- pixel values
- learning algorithm
- loss function
- semi supervised learning
- learning tasks
- machine learning algorithms
- classification models
- conditional random fields
- machine learning models
- learning problems
- genetic algorithm
- data mining
- classification accuracy
- feature vectors
- image processing
- data sets