Infrared and visible image fusion technologies are used to characterize the same scene using diverse modalities. However, most existing deep learning-based fusion methods are designed as symmetric networks, which ignore the differences between modal images and lead to source image information loss during feature extraction. In this paper, we propose a new fusion framework for the different characteristics of infrared and visible images. Specifically, we design a dual-stream asymmetric network with two different feature extraction networks to extract infrared and visible feature maps, respectively. The transformer architecture is introduced in the infrared feature extraction branch, which can force the network to focus on the local features of infrared images while still obtaining their contextual information. The visible feature extraction branch uses residual dense blocks to fully extract the rich background and texture detail information of visible images. In this way, it can provide better infrared targets and visible details for the fused image. Experimental results on multiple datasets indicate that DSA-Net outperforms state-of-the-art methods in both qualitative and quantitative evaluations. In addition, we also apply the fusion results to the target detection task, which indirectly demonstrates the fusion performances of our method.
Infrared and visible image fusion technologies are used to characterize the same scene by diverse modalities. However, most existing deep learning-based fusion methods are designed as symmetric networks, which ignore the differences between modal images and lead to the source image information loss during feature extraction. In this paper, we propose a new fusion framework for the different characteristics of infrared and visible images. Specifically, we design a dual-stream asymmetric network with two different feature extraction networks to extract infrared and visible feature maps respectively. The transformer architecture is introduced in the infrared feature extraction branch, which can force the network to focus on the local features of infrared images while still obtaining their contextual information. And the visible feature extraction branch uses residual dense blocks to fully extract the rich background and texture detail information of visible images. In this way, it can provide better infrared targets and visible details for the fused image. Experimental results on multiple datasets indicate that DSA-Net outperforms state-of-the-art methods in both qualitative and quantitative evaluations. In addition, we also apply the fusion results to the target detection task, which indirectly demonstrates the fusion performances of our method.
No abstract
Washington law allows people with criminal records that meet certain conditions to vacate their records, avoiding the harmful collateral consequences that accompany having a record. To estimate the size of the "second chance gap"-the share of individuals that could but haven't yet "expunged" their records, we modeled the eligibility criteria for vacatur and applied it to a sample of records obtained from the Administrative Office of the Courts of Washington. Importantly, data limitations made it impossible for us to consider out of state charges, payment of fines and fees, and definitive sentence completion, so we did not model them. Based on our analysis, we found that 60% of those who live burdened with criminal conviction records, or as many as 1M+ Washingtoninans, are potentially eligible to receive relief. But less than 3% of individuals eligible for relief, and less than 1% of the charges eligible for relief have received the remedy. At current rates of vacation, we estimate that it would take over 4,000 years to clear the backlog of charges alone, based on the gap and the actual number of charges that were vacated last year (1,973). Existing racial disparities in the Washington criminal justice system are significant: currently, African-Americans represent 4.2 % of individuals in Washington but within our sample, 11% of Washingtonians with a criminal record, 15% of Washingtonians with any felony record, and 22% of Washingtonians with a Class A felony record. We find that Clean Slate would reduce racial disparities among individuals in the general population while not necessarily worsening it among the population of people with records. Because of the large second chance gap, the filing of petitions by all those who are entitled to relief could result in a severe congestion at the courts. Washington can close the 97-99% second chance gap between eligibility and delivery of relief by automating relief, solving both problems, but only if it implements the law with some adjustments and compensates for missing and dirty data.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.