Gait recognition is a key biometric for long-distance identification,yet its performance is severely degraded by real-world challenges such as varying clothing,carrying conditions,and changing viewpoints.While combini...Gait recognition is a key biometric for long-distance identification,yet its performance is severely degraded by real-world challenges such as varying clothing,carrying conditions,and changing viewpoints.While combining silhouette and skeleton data is a promising direction,effectively fusing these heterogeneous modalities and adaptively weighting their contributions in response to diverse conditions remains a central problem.This paper introduces GaitMAFF,a novelMulti-modal Adaptive Feature Fusion Network,to address this challenge.Our approach first transforms discrete skeleton joints into a dense SkeletonMap representation to align with silhouettes,then employs an attention-based module to dynamically learn the fusion weights between the two modalities.These fused features are processed by a powerful spatio-temporal backbone withWeighted Global-Local Feature FusionModules(WFFM)to learn a discriminative representation.Extensive experiments on the challenging CCPG and Gait3D datasets show that GaitMAFF achieves state-of-the-art performance,with an average Rank-1 accuracy of 84.6%on CCPG and 58.7%on Gait3D.These results demonstrate that our adaptive fusion strategy effectively integrates complementary multimodal information,significantly enhancing gait recognition robustness and accuracy in complex scenes and providing a practical solution for real-world applications.展开更多
New high throughput DNA technologies resulted in a disproportion between the high number of scored markers for the mapping populations and relatively small sizes of the genotyped populations. Correspondingly, the numb...New high throughput DNA technologies resulted in a disproportion between the high number of scored markers for the mapping populations and relatively small sizes of the genotyped populations. Correspondingly, the number of markers may, by orders of magnitude, exceed the threshold of recombination resolution achievable for a given population size. Hence, only a small part of markers can be genuinely ordered in the map. The question is how to choose the most informative markers for building such a reliable “skeleton” map. We believe that our approach provides a solution to this difficult problem due to: a) powerful tools of discrete optimization for multilocus ordering;b) a verification procedure, which is impossible without fast and high-quality optimization, to control the map quality based on re-sampling techniques;c) an interactive algorithm of marker clustering in complicated situations caused by significant deviation of recombination rates between markers of non-homologous chromosomes from the expected 50% (referred to as quasi-linkage or pseudo-linkage);and d) an algorithm for detection and removing excessive markers to increase the stability of multilocus ordering.展开更多
基金funded by the Natural Science Foundation of Chongqing Municipality,grant number CSTB2022NSCQ-MSX0503.
文摘Gait recognition is a key biometric for long-distance identification,yet its performance is severely degraded by real-world challenges such as varying clothing,carrying conditions,and changing viewpoints.While combining silhouette and skeleton data is a promising direction,effectively fusing these heterogeneous modalities and adaptively weighting their contributions in response to diverse conditions remains a central problem.This paper introduces GaitMAFF,a novelMulti-modal Adaptive Feature Fusion Network,to address this challenge.Our approach first transforms discrete skeleton joints into a dense SkeletonMap representation to align with silhouettes,then employs an attention-based module to dynamically learn the fusion weights between the two modalities.These fused features are processed by a powerful spatio-temporal backbone withWeighted Global-Local Feature FusionModules(WFFM)to learn a discriminative representation.Extensive experiments on the challenging CCPG and Gait3D datasets show that GaitMAFF achieves state-of-the-art performance,with an average Rank-1 accuracy of 84.6%on CCPG and 58.7%on Gait3D.These results demonstrate that our adaptive fusion strategy effectively integrates complementary multimodal information,significantly enhancing gait recognition robustness and accuracy in complex scenes and providing a practical solution for real-world applications.
文摘New high throughput DNA technologies resulted in a disproportion between the high number of scored markers for the mapping populations and relatively small sizes of the genotyped populations. Correspondingly, the number of markers may, by orders of magnitude, exceed the threshold of recombination resolution achievable for a given population size. Hence, only a small part of markers can be genuinely ordered in the map. The question is how to choose the most informative markers for building such a reliable “skeleton” map. We believe that our approach provides a solution to this difficult problem due to: a) powerful tools of discrete optimization for multilocus ordering;b) a verification procedure, which is impossible without fast and high-quality optimization, to control the map quality based on re-sampling techniques;c) an interactive algorithm of marker clustering in complicated situations caused by significant deviation of recombination rates between markers of non-homologous chromosomes from the expected 50% (referred to as quasi-linkage or pseudo-linkage);and d) an algorithm for detection and removing excessive markers to increase the stability of multilocus ordering.
