Towards Learning Object Detectors with Limited Data for Industrial Applications / Karim Guirguis.

Format:

Book

Author/Creator:

Guirguis, Karim, author.

Language:

English

Subjects (All):

Artificial intelligence--Industrial applications--Congresses.

Artificial intelligence.

Physical Description:

1 online resource

Place of Publication:

[Place of publication not identified] : KIT Scientific Publishing, 2025.

Summary:

In this dissertation, three novel Generalized Few-Shot Object Detection (G-FSOD) approaches are presented to minimize the forgetting of previously learned classes while learning new classes with limited data. The first two approaches reduce the forgetting of base classes if they are still available during training. The third approach, for scenarios without base data, uses knowledge distillation to improve the knowledge transfer.

Notes:

Description based on publisher supplied metadata and other sources.