Real-time Instance Detection with Fast Incremental Learning

Object instance detection is a highly relevant task to several robotic applications such as automated order picking, or household and hospital assistance robots. In these applications, a holistic scene labeling is often not required whereas it is sufficient to find a certain object type of interest, e. g. for picking it up. At the same time, large and continuously changing object sets are characteristic in such applications, requiring efficient model update capabilities from the object detector. Today’s monolithic multi-class detectors do not fulfill this criterion for fast and flexible model updates. This paper introduces InstanceNet, an ensemble of efficient single-class instance detectors capable of fast and incremental adaptation to new object sets. Due to a dynamic sampling-based training strategy, accurate detection models for new objects can be obtained within less than 40 minutes on a consumer GPU while only a small percentage of the existing detection models needs to be updated in a very efficient manner. The new detector has been thoroughly evaluated on the basis of a novel dataset of 100 grocery store objects.

Authors

• Richard Bormann
• Xinjie Wang
• Markus Völk
• Kilian Kleeberger
• Jochen Lindermayr

Keywords

• Training
• Adaptation models
• Hospitals
• Conferences
• Graphics processing units
• Detectors
• Real-time systems
• Incremental Learning
• Fast Incremental Learning
• Object Detection
• Updated Model
• Object Instances
• Training Set
• Convolutional Neural Network
• Convolutional Layers
• Negative Samples
• Detection Performance
• ImageNet
• Bounding Box
• Batch Normalization
• Faster Convergence
• RGB Images
• Similar Appearance
• Training Efficiency
• Faster R-CNN
• Metric Learning
• Fast Training
• Few-shot Learning
• Bounding Box Regression
• Beginning Of Training
• Single Shot Detector
• Update Time
• Training Data
• Object Dataset
• Feature Points
• Validation Set
• Classification Loss