Self-Guided Proposal Generation for Weakly Supervised Object Detection

Cheng, Gong; Xie, Xuan; Chen, Wei-Ning; Feng, Xiao; Yao, Xiwen; Han, Junwei

doi:10.1109/tgrs.2022.3181466

Cited by 24 publications

(15 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is a significant improvement because MIL-based approaches cannot produce high-quality detectors starting from low-quality proposals, as already observed in [41]. This is further pointed out by Cheng et al [51] who propose SPG based on an RPN that exploits the objectness confidence score to generate high-quality proposals. The authors show that using the proposed RPN in place of standard techniques (e.g., Selective Search) can improve the performance of previous MIL-based methods such as OICR [14] and MELM [67].…”

Section: Mil-basedmentioning

confidence: 85%

“…It has been demonstrated that these proposals cannot cover the entire object well, severely hindering the performance of WSOD. This problem is effectively addressed by Cheng et al [51] through a region proposal network. Using the RPN with the Min-Entropy Latent Model (MELM) [67], the authors can obtain an improvement of 7.11% over the basic MELM method [67], confirming the importance of high-quality proposals.…”

Section: Dior Datasetmentioning

confidence: 99%

“…Better initial proposals: As shown by Cheng et al [51], proposal generation is a very critical step because the performance of an MIL-based method is strongly dependent on the quality of the initial proposal. For this reason, developing novel proposal generation methods is fundamental, especially when specific use-cases are addressed, and could boost the performance of WS approaches.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Weakly Supervised Object Detection for Remote Sensing Images: A Survey

et al. 2022

View full text Add to dashboard Cite

The rapid development of remote sensing technologies and the availability of many satellite and aerial sensors have boosted the collection of large volumes of high-resolution images, promoting progress in a wide range of applications. As a consequence, Object detection (OD) in aerial images has gained much interest in the last few years. However, the development of object detectors requires a massive amount of carefully labeled data. Since annotating datasets is very time-consuming and may require expert knowledge, a consistent number of weakly supervised object localization (WSOL) and detection (WSOD) methods have been developed. These approaches exploit only coarse-grained metadata, typically whole image labels, to train object detectors. However, many challenges remain open due to the missing location information in the training process of WSOD approaches and to the complexity of remote sensing images. Furthermore, methods studied for natural images may not be directly applicable to remote sensing images (RSI) and may require carefully designed adaptations. This work provides a comprehensive survey of the recent achievements of remote sensing weakly supervised object detection (RSWSOD). An analysis of the challenges related to RSWSOD is presented, the advanced techniques developed to improve WSOD are summarized, the available benchmarking datasets are described and a discussion of future directions of RSWSOD research is provided.

show abstract

Section: Mil-basedmentioning

confidence: 85%

Section: Dior Datasetmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Weakly Supervised Object Detection for Remote Sensing Images: A Survey

et al. 2022

View full text Add to dashboard Cite

show abstract

“…An IoU-balanced sampling component was constructed to excavate the complete instances, which further enhanced the performance of WSOD. Cheng et al [23] proposed a self-guided proposal generation (SPG) module to excavate more reliable candidate proposals, which explicitly generated more high-quality instance-level features. Tan et al [37] unified horizontal and oriented object detection tasks into a complete WSOD framework to detect different oriented instances in RSIs.…”

Section: Instance-level Feature Refinementmentioning

confidence: 99%

“…According to [21][22][23][24][25][26][27][28], WSOD in RSIs still encounters two major challenges. First of all, most previous methods tend to focus on the most discriminative regions in an image (part domination).…”

Section: Introductionmentioning

confidence: 99%

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

Zheng,

Wu,

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Weakly supervised object detection (WSOD) aims to predict a set of bounding boxes and corresponding category labels for instances with only image-level supervisions. Compared with fully supervised object detection, WSOD in remote sensing images (RSIs) is much more challenging due to the vast foreground-related context regions. In this paper, we propose a progressive image-level and instance-level feature refinement network to address the problems of missing detection and part domination for WSOD in RSIs. Firstly, we propose a multi-label attention mining loss (MAML)-guided image-level feature refinement branch to effectively allocate the computational resources towards the most informative part of images. With the supervision of MAML, all latent instances in images are emphasized. However, image-level feature refinement further expands responsive gaps between the informative part and other sub-optimal informative ones, which results in exacerbating the problem of part domination. In order to alleviate the above-mentioned limitation, we further construct an instance-level feature refinement branch to re-balance the contributions of different adjacent candidate bounding boxes according to the detection task. An instance selection loss (ISL) is proposed to progressively boost the representation of salient regions by exploring supervision from the network itself. Finally, we integrate the image-level and instance-level feature refinement branches into a complete network and the proposed MAML and ISL functions are merged with class classification and box regression to optimize the whole WSOD network in an end-to-end training fashion. We conduct experiments on two popular WSOD datasets, NWPU VHR-10.v2 and DIOR. All the experimental results demonstrate that our method achieves a competitive performance compared with other state-of-the-art approaches.

show abstract