tailieunhanh - LabelMe: a database and web-based tool for image annotation

In the Arab Region, digital migration is still in its very early years, but nevertheless is being driven by the high proportion of young demographic in many Arab countries. With 55% of the Arab Region population under the age of 252 , this segment is expected to drive the growth of digital media. Since international players have been experimenting with online content and business models for some time, the Arab Region now has the opportunity to learn from the mistakes and successes of the European, North American and Asian media markets to really drive growth | International Journal of Computer Vision Volume 77 Issue 1-3 PAGES 157-173 MAY 2008 LabelMe a database and web-based tool for image annotation Bryan c. Russell Antonio Torralba Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge MA 02139 USA brussell@ torralba@ Kevin P Murphy Departments of computer science and statistics University of British Columbia Vancouver BC V6T1Z4 murphyk@c William T. Freeman Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge MA 02139 USA billf@ Abstract We seek to build a large collection of images with ground truth labels to be used for object detection and recognition research. Such data is useful for supervised learning and quantitative The first two authors contributed equally to this work. 0 evaluation. To achieve this we developed a web-based tool that allows easy image annotation and instant sharing of such annotations. Using this annotation tool we have collected a large dataset that spans many object categories often containing multiple instances over a wide variety of images. We quantify the contents of the dataset and compare against existing state of the art datasets used for object recognition and detection. Also we show how to extend the dataset to automatically enhance object labels with WordNet discover object parts recover a depth ordering of objects in a scene and increase the number of labels using minimal user supervision and images from the web. 1 Introduction Thousands of objects occupy the visual world in which we live. Biederman 4 estimates that humans can recognize about 30000 entry-level object categories. Recent work in computer vision has shown impressive results for the detection and recognition of a few different object categories 42 16 22 . However the size and contents of existing datasets among other factors limit current methods from scaling to thousands .

• Cơ sở dữ liệu
• Intimate Relationship
• Stomatal Responses
• Stress and Air Humidity
• Genetics and Adaptation
• Plant Genes
• Abiotic Stress
• Abiotic Stres