Abstract:Weakly supervised object detection (WSOD) in remote sensing images (RSIs) has attracted more and more attention because its training merely relies on image-level category labels, which significantly reduces the cost of manual annotation. With the exploration of WSOD, it has obtained many promising results. However, most of the WSOD methods still have two challenges. The first challenge is that the detection results of WSOD tend to locate the significant regions of the object but not the overall object. The second challenge is that the traditional pseudo-instance label assignment strategy cannot adapt to the quality distribution change of proposals during training, which is not conducive to training a high-performance detector. To tackle the first challenge, a novel high-quality seed instance mining (HSIM) module is designed to mine high-quality seed instances. Specifically, the proposal comprehensive score (PCS) that consists of the traditional proposal score (PS) and the proposal space contribution score (PSCS) is designed as a novel metric to mine seed instances, where the PS indicates the probability that a proposal pertains to a certain category and the PSCS is calculated by the spatial correlation between top-scoring proposals, which is utilized to evaluate the wholeness with which a proposal locates an object. Consequently, the high PCS will encourage the WSOD model to mine the high-quality seed instances. To tackle the second challenge, a dynamic pseudo-instance label assignment (DPILA) strategy is developed by dynamically setting the label assignment threshold to train high-quality instances. Consequently, the DPILA can better adapt the distribution change of proposals according to the dynamic threshold during training and further promote model performance. The ablation studies verify the validity of the proposed PCS and DPILA. The comparison experiments verify that our method obtains better performance than other advanced WSOD methods on two popular RSIs datasets.

What problem does this paper attempt to address?

This paper attempts to address two major challenges in weakly - supervised object detection (WSOD) in remote sensing images: 1. **Detection results tend to locate the salient regions of objects rather than the whole objects**: Most existing WSOD methods only use the Proposal Score (PS) to mine seed instances. This method often leads to detection results that are only concentrated on the salient parts of the object, rather than the entire object. This is particularly disadvantageous in remote sensing images with high background noise, because these methods may miss the overall localization of the object. 2. **Traditional pseudo - instance label assignment strategies cannot adapt to the changes in the proposal quality distribution during the training process**: Traditional methods usually set a fixed label assignment threshold (such as the IoU value) to determine whether a proposal is a positive or negative example. However, as the training progresses, the quality distribution of the proposals will change, and the fixed threshold may no longer be applicable, which is not conducive to training high - quality detectors. To address these two challenges, the authors propose the following solutions: - **High - Quality Seed Instance Mining (HSIM) module**: A new Proposal Comprehensive Score (PCS) is designed, which combines the traditional Proposal Score (PS) and the Proposal Space Contribution Score (PSCS). PSCS measures the degree of complete localization of the object by the proposal by considering the spatial relationships between high - scoring proposals. In this way, the HSIM module can more accurately mine high - quality seed instances, not just the salient regions. - **Dynamic Pseudo - Instance Label Assignment (DPILA) strategy**: A method for dynamically adjusting the label assignment threshold is developed, so that the threshold can change as the training progresses. Specifically, the DPILA strategy dynamically calculates the label assignment threshold by carefully designing a function that increases with the number of iterations, so as to better adapt to the changes in the proposal quality distribution and increase the number of positive examples in the early stage of training, further improving the performance of the model. Through these innovations, the paper aims to improve the accuracy and robustness of weakly - supervised object detection in remote sensing images.

High-Quality Instance Mining and Dynamic Label Assignment for Weakly Supervised Object Detection in Remote Sensing Images

High-Quality Instance Mining and Weight Re-Assigning for Weakly Supervised Object Detection in Remote Sensing Images

SAM-Induced Pseudo Fully Supervised Learning for Weakly Supervised Object Detection in Remote Sensing Images

Spatial Likelihood Voting with Self-Knowledge Distillation for Weakly Supervised Object Detection.

Multiscale Image Splitting Based Feature Enhancement and Instance Difficulty Aware Training for Weakly Supervised Object Detection in Remote Sensing Images

Weakly Supervised Object Detection for Remote Sensing Images via Progressive Image-Level and Instance-Level Feature Refinement

SLV: Spatial Likelihood Voting for Weakly Supervised Object Detection

MOL: Towards Accurate Weakly Supervised Remote Sensing Object Detection Via Multi-view Noisy Learning

Self-Training-Based Semantic-Balanced Network for Weakly Supervised Object Detection in Remote-Sensing Images

Progressive Contextual Instance Refinement for Weakly Supervised Object Detection in Remote Sensing Images

Attention Erasing and Instance Sampling for Weakly Supervised Object Detection

Refining and reweighting pseudo labels for weakly supervised object detection

Point-Based Weakly Supervised Learning for Object Detection in High Spatial Resolution Remote Sensing Images

Localizing From Classification: Self-Directed Weakly Supervised Object Localization for Remote Sensing Images

Robust Instance-Based Semi-Supervised Learning Change Detection for Remote Sensing Images

Self-Guided Proposal Generation for Weakly Supervised Object Detection

Object Detection in Optical Remote Sensing Images Based on Weakly Supervised Learning and High-Level Feature Learning

Dynamic Pseudo-Label Generation for Weakly Supervised Object Detection in Remote Sensing Images

Complete Instances Mining for Weakly Supervised Instance Segmentation

Instance Mining with Class Feature Banks for Weakly Supervised Object Detection.

Instance-Level Contrastive Learning for Weakly Supervised Object Detection