Abstract:With the rapid development of remote sensing technology and the growing demand for applications, the classical deep learning-based object detection model is bottlenecked in processing incremental data, especially in the increasing classes of detected objects. It requires models to sequentially learn new classes of objects based on the current model, while preserving old categories-related knowledge. Existing class-incremental detection methods achieve this goal mainly by constraining the optimization trajectory in the feature of output space. However, these works neglect the case where the previously learned background is a new category to learn, resulting in performance degradation in the new category because of the conflict between remaining the background-related knowledge or updating the background-related knowledge. This paper proposes a novel class-incremental detection method incorporated with the teacher-student architecture and the selective distillation (SDCID) strategy. Specifically, it is the asymmetry architecture, i.e., the teacher network temporarily stores historical knowledge of previously learned objects, and the student network integrates historical knowledge from the teacher network with the newly learned object-related knowledge, respectively. This asymmetry architecture reveals the significance of the distinct representation of history knowledge and new knowledge in incremental detection. Furthermore, SDCID selectively masks the shared subobject of new images to learn and previously learned background, while learning new categories of images and then transfers the classification results of the student model to the background class following the judgment model of the teacher model. This manner avoids interferences between the background category-related knowledge from a teacher model and the learning of other new classes of objects. In addition, we proposed a new incremental learning evaluation metric, C-SP, to comprehensively evaluate the incremental learning stability and plasticity performance. We verified the proposed method on two object detection datasets of remote sensing images, i.e., DIOR and DOTA. The experience results in accuracy and C-SP suggest that the proposed method surpasses existing class-incremental detection methods. We further analyzed the influence of the mask component in our method and the hyper-parameters sensitive to our method.

Multi-View correlation distillation for incremental object detection

Multi-View Domain Adaptive Object Detection on Camera Networks.

RD-IOD: Two-Level Residual-Distillation-Based Triple-Network for Incremental Object Detection

Enhancing class-incremental object detection in remote sensing through instance-aware distillation

Incremental Detection of Remote Sensing Objects with Feature Pyramid and Knowledge Distillation

Class incremental learning of remote sensing images based on class similarity distillation

A New Multinetwork Mean Distillation Loss Function for Open-World Domain Incremental Object Detection

A Class-Incremental Detection Method of Remote Sensing Images Based on Selective Distillation

Response-based Distillation for Incremental Object Detection

Multi-Task Incremental Learning for Object Detection

Towards Non Co-occurrence Incremental Object Detection with Unlabeled In-the-Wild Data

Overcoming Catastrophic Forgetting in Incremental Object Detection via Elastic Response Distillation

Feature Estimations Based Correlation Distillation for Incremental Image Retrieval

Incremental Learning of Object Detectors without Catastrophic Forgetting

SID: Incremental learning for anchor-free object detection via Selective and Inter-related Distillation

Class-Incremental SAR Obeject Detection Via Adaptive Distributed Response Distillation

Bridging Non Co-occurrence with Unlabeled In-the-wild Data for Incremental Object Detection

Bridge Past and Future: Overcoming Information Asymmetry in Incremental Object Detection

M2KD: Multi-model and Multi-level Knowledge Distillation for Incremental Learning

Incremental Object Detection with Image-level Labels

Purified Distillation: Bridging Domain Shift and Category Gap in Incremental Object Detection