scispace - formally typeset
Search or ask a question
Topic

Object (computer science)

About: Object (computer science) is a research topic. Over the lifetime, 106024 publications have been published within this topic receiving 1360115 citations. The topic is also known as: obj & Rq.


Papers
More filters
Posted Content
TL;DR: This work proposes a novel solution for semi-supervised video object segmentation by leveraging memory networks and learning to read relevant information from all available sources to better handle the challenges such as appearance changes and occlussions.
Abstract: We propose a novel solution for semi-supervised video object segmentation. By the nature of the problem, available cues (e.g. video frame(s) with object masks) become richer with the intermediate predictions. However, the existing methods are unable to fully exploit this rich source of information. We resolve the issue by leveraging memory networks and learn to read relevant information from all available sources. In our framework, the past frames with object masks form an external memory, and the current frame as the query is segmented using the mask information in the memory. Specifically, the query and the memory are densely matched in the feature space, covering all the space-time pixel locations in a feed-forward fashion. Contrast to the previous approaches, the abundant use of the guidance information allows us to better handle the challenges such as appearance changes and occlussions. We validate our method on the latest benchmark sets and achieved the state-of-the-art performance (overall score of 79.4 on Youtube-VOS val set, J of 88.7 and 79.2 on DAVIS 2016/2017 val set respectively) while having a fast runtime (0.16 second/frame on DAVIS 2016 val set).

214 citations

Patent
03 Apr 1991
TL;DR: In this article, an object based data processing system including an extensible set of object types and a corresponding set of "object managers" wherein each object manager is a program for operating with the data stored in a corresponding type of object.
Abstract: An object based data processing system including an extensible set of object types and a corresponding set of "object managers" wherein each object manager is a program for operating with the data stored in a corresponding type of object. The object managers in general support at least a standard set of operations. Any program can effect performance of these standard operations on objects of any type by making an "invocation" request. In response to an invocation request, object management services (which are available to all object managers) identifies and invokes an object manager that is suitable for performing the requested operation on the specified type of data. A mechanism is provided for linking data from one object into another object. An object catalog includes both information about objects and about links between objects. Data interchange services are provided for communicating data between objects of different types, using a set of standard data interchange formats. A matchmaker facility permits two processes that are to cooperate in a data interchange operation identify each other and to identify data formats they have in common. A facility is provided for managing shared data "resources". Customized versions of resources can be created and co-exist with standard resources. A resource retrieval function determines whether a customized or a standard resource is to be returned in response to each request for a resource.

213 citations

Posted Content
TL;DR: The GuessWhat?! dataset as mentioned in this paper is a large-scale dataset consisting of 150k human-played games with a total of 800k visual question-answer pairs on 66k images.
Abstract: We introduce GuessWhat?!, a two-player guessing game as a testbed for research on the interplay of computer vision and dialogue systems. The goal of the game is to locate an unknown object in a rich image scene by asking a sequence of questions. Higher-level image understanding, like spatial reasoning and language grounding, is required to solve the proposed task. Our key contribution is the collection of a large-scale dataset consisting of 150K human-played games with a total of 800K visual question-answer pairs on 66K images. We explain our design decisions in collecting the dataset and introduce the oracle and questioner tasks that are associated with the two players of the game. We prototyped deep learning models to establish initial baselines of the introduced tasks.

213 citations

Patent
02 Jan 2001
TL;DR: A data package containing instructions for manipulating user data as discussed by the authors is a data package that includes a header identifying the data package and containing information on content in the package; transaction objects operable to effect a change to user data; and object operation instructions.
Abstract: A data package containing instructions for manipulating user data. The package includes a header identifying the data package and containing information on content in the data package; transaction objects operable to effect a change to user data; and object operation instructions. The objects can be binary data or structured data. A grouping of data packages can contain changes for a user's data across multiple systems.

213 citations


Network Information
Related Topics (5)
Query optimization
17.6K papers, 474.4K citations
84% related
Programming paradigm
18.7K papers, 467.9K citations
84% related
Software development
73.8K papers, 1.4M citations
83% related
Compiler
26.3K papers, 578.5K citations
83% related
Software system
50.7K papers, 935K citations
82% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
202238
20213,087
20205,900
20196,540
20185,940
20175,046