English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

The benefits of deep neural networks (DNNs) have become of interest forsafety critical applications like medical ones or automated driving. Here,however, quantitative insights into the DNN inner representations aremandatory. One approach to this is concept analysis, which aims to establish amapping between the internal representation of a DNN and intuitive semanticconcepts. Such can be sub-objects like human body parts that are valuable forvalidation of pedestrian detection. To our knowledge, concept analysis has notyet been applied to large object detectors, specifically not for sub-parts.Therefore, this work first suggests a substantially improved version of theNet2Vec approach (arXiv:1801.03454) for post-hoc segmentation of sub-objects.Its practical applicability is then demonstrated on a new concept dataset bytwo exemplary assessments of three standard networks, including the larger MaskR-CNN model (arXiv:1703.06870): (1) the consistency of body part similarity,and (2) the invariance of internal representations of body parts with respectto the size in pixels of the depicted person. The findings show that therepresentation of body parts is mostly size invariant, which may suggest anearly intelligent fusion of information in different size categories.

Verification of Size Invariance in DNN Activations using Concept Embeddings