Abstract-A novel sub-Markov random walk (subRW) algorithm with label prior is proposed for seeded image segmentation, which can be interpreted as a traditional random walker on a graph with added auxiliary nodes. Under this explanation, we unify the proposed subRW and other popular random walk (RW) algorithms. This unifying view will make it possible for transferring intrinsic findings between different RW algorithms, and offer new ideas for designing novel RW algorithms by adding or changing auxiliary nodes. To verify the second benefit, we design a new subRW algorithm with label prior to solve the segmentation problem of objects with thin and elongated parts. The experimental results on both synthetic and natural images with twigs demonstrate that the proposed subRW method outperforms previous RW algorithms for seeded image segmentation. [20] for general image segmentation. In their work, the user should give labels to a small number of pixels. Assuming a random walker starts at each unlabeled pixel, they calculate the probabilities that these random walkers first reach the pre-labeled pixels. These probabilities will determine the probability of each unlabeled pixel belonging to a label. By assigning each pixel to a label with the greatest probability, the interactive image segmentation Manuscript received August 20, 2015; revised November 16, 2015; accepted November 28, 2015 [17] have been proposed, the RW has been extended to segment out disconnected objects by using prior models without labeling each object. In other words, a user only needs to indicate labels on some objects and the other similar objects will be segmented out. Sinop and Grady [23] proposed a common framework to unify the RW, graph cuts, and shortest path algorithms for interactive segmentation. Furthermore, they added the popular watershed segmentation algorithm to this framework [35] and made the theoretical analysis for the connections between these algorithms. This unified framework brings some benefits, including opening new possibilities for using unary terms in traditional watershed algorithms to optimize more general models. Index Terms-Recently, some researchers [27], [40] have focused on segmenting natural images with complex textures. They extend the RW algorithm and obtain better performance for these challenging images. In general, these algorithms are graphbased so we can use a graph to describe an image for introducing them. Kim et al. [27] proposed a random walker with a restarting probability (RWR) for segmentation. It means that this random walker will return to the starting node with a probability c at each step, and walk to other adjacent nodes with probability 1 − c. Shen et al. [40] have developed the lazy random walk (LRW) for superpixel segmentation. A LRW will stay at the current node with a probability 1 − α and walk out along the edges connected with the current node with probability α. Wu et al. [37] proposed another similar RW algorithm called partially absorbing random walk (PARW) for applications based...
Hyperparameters are numerical presets whose values are assigned prior to the commencement of the learning process. Selecting appropriate hyperparameters is critical for the accuracy of tracking algorithms, yet it is difficult to determine their optimal values, in particular, adaptive ones for each specific video sequence. Most hyperparameter optimization algorithms depend on searching a generic range and they are imposed blindly on all sequences. Here, we propose a novel hyperparameter optimization method that can find optimal hyperparameters for a given sequence using an action-prediction network leveraged on Continuous Deep Q-Learning. Since the common state-spaces for object tracking tasks are significantly more complex than the ones in traditional control problems, existing Continuous Deep Q-Learning algorithms cannot be directly applied. To overcome this challenge, we introduce an efficient heuristic to accelerate the convergence behavior. We evaluate our method on several tracking benchmarks and demonstrate its superior performance 1 .
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.