While significant progress has been made in multi-modal learning driven by large-scale image-text datasets, there is still a noticeable gap in the availability of such datasets within the facial domain. To facilitate ...
详细信息
Multivariate Time Series Classification (MTSC) enables the analysis if complex temporal data, and thus serves as a cornerstone in various real-world applications, ranging from healthcare to finance. Since the relation...
详细信息
Commit messages are natural language descriptions of code changes, which are important for software evolution such as code understanding and maintenance. However, previous methods are trained on the entire dataset wit...
详细信息
High-power continuous wave fiber laser cutting is a next-generation cutting technology for rapid prototyping and small-scale fabrication. To utilize the advantage offered by this technology in terms of high-quality cu...
详细信息
With increasing hardware computing power and model capacity, visual tasks for scene cognitive understanding have attracted more attention, such as visual relationships inference. The scene graph representation formed ...
详细信息
With increasing hardware computing power and model capacity, visual tasks for scene cognitive understanding have attracted more attention, such as visual relationships inference. The scene graph representation formed ...
详细信息
ISBN:
(纸本)9781450387835
With increasing hardware computing power and model capacity, visual tasks for scene cognitive understanding have attracted more attention, such as visual relationships inference. The scene graph representation formed by a coupling of objects, attributes and relationships nodes displayed by different modalities of information, including original image, foreground things, background stuff and scene attributes, strongly promotes the progress of research area. In this paper, we address the scene graph representation of traffic scenarios for autonomous driving. It should be noted that the universal representation are the specific needs of cognitive understanding of traffic scenes: on the one hand, there is a lack of fine-grained description of key objects and attributes; on the other hand, there are redundant descriptions of objects and relationships. To tackle these problems, we take advantage of the fine-grained instance-level annotation of the traffic scene, proposing a bottom-up representation paradigm. It makes full use of the hierarchical structure of the traffic scene and the sparsity of element classes. In addition, on the basis of the existing methods, we optimize the relationship list of traffic scene graph representation. Moreover, we improve the scene graph annotation methods, proposing a "ground-vision joint location method" to better describe the spatially-distributed visual knowledge. The case analysis showed that compared with existing methods, our paradigm for scene graph can represent more abundant traffic scene information.
With the rapid development of artificial intelligence (AI) in medical image processing, deep learning in color fundus photography (CFP) analysis is also evolving. Although there are some open-source, labeled datasets ...
详细信息
Person search is a challenging task due to the different requirements of annotations between person detection and Re-identification. In general, person search methods use the supervised person Re-identification method...
详细信息
Person search is a challenging task due to the different requirements of annotations between person detection and Re-identification. In general, person search methods use the supervised person Re-identification methods, where abundant identity labels of the bounding boxes are essential. However, most person images are unlabeled in the real-world scenario and it is unpractical to annotate the abundant fine-grained labels for unlabeled images. Obviously, the existing supervised methods are not appropriate with the real-world scenario. Therefore, we propose an unsupervised learning method for person search in this paper, which contacts two parts: one is unsupervised person detection and the other is unsupervised person Re-identification. The experimental results on two well-known datasets, CUHK-SYSU and PRW, indicate that proposed method achieves competitive performance than the state-of-art unsupervised methods. Note that proposed method has greater practical significance even though it does not get the results as good as the general supervised methods.
The Association for the Advancement of Artificial Intelligence (AAAI) presented the 2010 Fall Symposium Series on November 11-13, 2010. The eight symposia included Cognitive and Metacognitive Educational Systems, Comm...
详细信息
The Association for the Advancement of Artificial Intelligence (AAAI) presented the 2010 Fall Symposium Series on November 11-13, 2010. The eight symposia included Cognitive and Metacognitive Educational Systems, Commonsense Knowledge, Complex Adaptive Systems: Resilience, Robustness, and Evolvability, Computational Models of Narrative, Dialog with Robots, Manifold Learning and Its Applications, Proactive Assistant Agents and Quantum Informatics for Cognitive, Social, and Semantic Processes. Cognitive and Metacognitive Educational Systems aimed to provide a comprehensive definition of metacognitive educational systems that is inclusive of the theoretical, architectural, and educational aspects of this field. The AAAI Commonsense Knowledge Fall Symposium had the goal of bringing together the diverse elements of this community whose work benefits from or contributes to the representation of general knowledge about the world. One of the specific goals of Proactive symposium was to gather the researchers from various projects in assistant agents to share their wisdom in retrospect.
In this paper we are proposing an information integration approach based on minimalist-upper ontology that can be applied among Web 2.0 applications. Instead of following conventional ontology engineering principles w...
详细信息
暂无评论