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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The human vision and perception system is inherently incremental where newknowledge is continually learned over time whilst existing knowledge isretained. On the other hand, deep learning networks are ill-equipped forincremental learning. When a well-trained network is adapted to new categories,its performance on the old categories will dramatically degrade. To addressthis problem, incremental learning methods have been explored which preservethe old knowledge of deep learning models. However, the state-of-the-artincremental object detector employs an external fixed region proposal methodthat increases overall computation time and reduces accuracy comparing toRegion Proposal Network (RPN) based object detectors such as Faster RCNN. Thepurpose of this paper is to design an efficient end-to-end incremental objectdetector using knowledge distillation. We first evaluate and analyze theperformance of the RPN-based detector with classic distillation on incrementaldetection tasks. Then, we introduce multi-network adaptive distillation thatproperly retains knowledge from the old categories when fine-tuning the modelfor new task. Experiments on the benchmark datasets, PASCAL VOC and COCO,demonstrate that the proposed incremental detector based on Faster RCNN is moreaccurate as well as being 13 times faster than the baseline detector.

Faster ILOD: Incremental Learning for Object Detectors based on Faster RCNN