Graph Neural Networks (GNNs) have gained popularity in image matching methods, proving useful for various computer vision tasks like Structure from Motion (SfM) and 3D reconstruction. A well-known example is SuperGlue. Lightweight variants, such as LightGlue, have been developed with a focus on stacking fewer GNN layers compared to SuperGlue. This paper proposes the h-GNN, a lightweight image matching model, with improvements in the two processing modules, the GNN and matching modules. After image features are detected and described as keypoint nodes of a base graph, the GNN module, which primarily aims at increasing the h-GNN’s depth, creates successive hierarchies of compressed-size graphs from the base graph through a clustering technique termed SC+PCA. SC+PCA combines Principal Component Analysis (PCA) with Spectral Clustering (SC) to enrich nodes with local and global information during graph clustering. A dual non-contrastive clustering loss is used to optimize graph clustering. Additionally, four message-passing mechanisms have been proposed to only update node representations within a graph cluster at the same hierarchical level or to update node representations across graph clusters at different hierarchical levels. The matching module performs iterative pairwise matching on the enriched node representations to obtain a scoring matrix. This matrix comprises scores indicating potential correct matches between the image keypoint nodes. The score matrix is refined with a ‘dustbin’ to further suppress unmatched features. There is a reprojection loss used to optimize keypoint match positions. The Sinkhorn algorithm generates a final partial assignment from the refined score matrix. Experimental results demonstrate the performance of the proposed h-GNN against competing state-of-the-art (SOTA) GNN-based methods on several image matching tasks under homography, estimation, indoor and outdoor camera pose estimation, and 3D reconstruction on multiple datasets. Experiments also demonstrate improved computational memory and runtime, approximately 38.1% and 26.14% lower than SuperGlue, and an average of about 6.8% and 7.1% lower than LightGlue. Future research will explore the effects of integrating more recent simplicial message-passing mechanisms, which concurrently update both node and edge representations, into our proposed model.
28 November 2022 in Asian Journal of Research in Computer Science
CCTV monitoring system is an essential security tool for visual surveillance accelerating the investigation in potential criminal activities when the need arises. Although expensive, universities mostly with public-access campuses in general, all need this system mainly to maintain safety and security in real-time, allowing legitimate students and staff to access campus resources and concurrently preventing any unauthorized persons access within the campus as well as responding to incidents with necessary action. C. K. Tedam University of Technology and Applied Sciences (CKT-UTAS) is a university in the Upper East Region of Ghana that does not have such a monitoring system. Since it is a newly established public university, its allocated funds are limited and could not be used to establish such an expensive system. To supplement their ongoing efforts in building security monitoring system, this study constitutes the blueprint procedures for building economical but reliable and efficient CCTV camera system for monitoring the property on campus and also the in and out of students and university staff members. The CCTV system was tested in monitoring vantage security post of the University. After observing and analyzing the trends in data from both the physical and our proposed automated monitoring approach, it can be concluded that the CCTV camera setup outperforms the physical and manual form or monitoring vantage security posts on University campus. Since the of monitoring security post using the proposed CCTV setup is advantageous in requiring lesser human effort and skills, it can be recommended for universities with low income-flow and low budget. This probably can make the university campus more secure and reliable.
Word embedding have been used in numerous Natural Language Processing and Machine Learning tasks. However, it is a high-dimensional vector field that propagate stereotypes to software applications. Its current debiasing frameworks do not completely capture its embedded patterns. In this paper, we propose deb2viz, a visual debiasing approach that explores and manipulates the high-dimensional patterns of word embedding field. First, we partition this vector field into interrelated low-dimensional subspaces to equalize and neutralize distances between gender-definitional and gender-neutral words. To further reduce gender bias, we update the distances of appropriate nearest neighbors for gender-neutral words to be arbitrarily close. Experimental results on several benchmark standards show the competitiveness of our proposed method in mitigating bias within pre-trained word2vec embedding model.