Verification of Size Invariance in DNN Activations Using Concept Embeddings
The benefits of deep neural networks (DNNs) have become of interest for safety critical applications like medical ones or automated driving. Here, however, quantitative insights into the DNN inner representations are mandatory [10]. One approach to this is concept analysis, which aims to establish a...
Saved in:
Published in | Artificial Intelligence Applications and Innovations pp. 374 - 386 |
---|---|
Main Author | |
Format | Book Chapter |
Language | English |
Published |
Cham
Springer International Publishing
|
Series | IFIP Advances in Information and Communication Technology |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The benefits of deep neural networks (DNNs) have become of interest for safety critical applications like medical ones or automated driving. Here, however, quantitative insights into the DNN inner representations are mandatory [10]. One approach to this is concept analysis, which aims to establish a mapping between the internal representation of a DNN and intuitive semantic concepts. Such can be sub-objects like human body parts that are valuable for validation of pedestrian detection. To our knowledge, concept analysis has not yet been applied to large object detectors, specifically not for sub-parts. Therefore, this work first suggests a substantially improved version of the Net2Vec approach [5] for post-hoc segmentation of sub-objects. Its practical applicability is then demonstrated on a new concept dataset by two exemplary assessments of three standard networks, including the larger Mask R-CNN model [9]: (1) the consistency of body part similarity, and (2) the invariance of internal representations of body parts with respect to the size in pixels of the depicted person. The findings show that the representation of body parts is mostly size invariant, which may suggest an early intelligent fusion of information in different size categories. |
---|---|
Bibliography: | The research leading to these results was partly funded by the German Federal Ministry for Economic Affairs and Energy within the project “KI-Absicherung”. |
ISBN: | 9783030791490 3030791491 |
ISSN: | 1868-4238 1868-422X |
DOI: | 10.1007/978-3-030-79150-6_30 |