The Experts below are selected from a list of 183 Experts worldwide ranked by ideXlab platform

A. Cioppa - One of the best experts on this subject based on the ideXlab platform.

  • A Bottom-Up Approach Based on Semantics for the Interpretation of the Main Camera Stream in Soccer Games
    2018 IEEE CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2018
    Co-Authors: A. Cioppa, A. Deliège, M. Van Droogenbroeck
    Abstract:

    Automatic interpretation of sports games is a major challenge, especially when these sports feature complex players organizations and game phases. This paper describes a bottom-up approach based on the extraction of semantic features from the video Stream of the main Camera in the particular case of soccer using scene-specific techniques. In our approach, all the features, ranging from the pixel level to the game event level, have a semantic meaning. First, we design our own scene-specific deep learning semantic segmentation network and hue histogram analysis to extract pixel-level semantics for the field, players, and lines. These pixel-level semantics are then processed to compute interpretative semantic features which represent characteristics of the game in the video Stream that are exploited to interpret soccer. For example, they correspond to how players are distributed in the image or the part of the field that is filmed. Finally, we show how these interpretative semantic features can be used to set up and train a semantic-based decision tree classifier for major game events with a restricted amount of training data. The main advantages of our semantic approach are that it only requires the video feed of the main Camera to extract the semantic features, with no need for Camera calibration, field homography, player tracking, or ball position estimation. While the automatic interpretation of sports games remains challenging, our approach allows us to achieve promising results for the semantic feature extraction and for the classification between major soccer game events such as attack, goal or goal opportunity, defense, and middle game.

  • CVPR Workshops - A Bottom-Up Approach Based on Semantics for the Interpretation of the Main Camera Stream in Soccer Games
    2018 IEEE CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2018
    Co-Authors: A. Cioppa, A. Deliège, M. Van Droogenbroeck
    Abstract:

    Automatic interpretation of sports games is a major challenge, especially when these sports feature complex players organizations and game phases. This paper describes a bottom-up approach based on the extraction of semantic features from the video Stream of the main Camera in the particular case of soccer using scene-specific techniques. In our approach, all the features, ranging from the pixel level to the game event level, have a semantic meaning. First, we design our own scene-specific deep learning semantic segmentation network and hue histogram analysis to extract pixel-level semantics for the field, players, and lines. These pixel-level semantics are then processed to compute interpretative semantic features which represent characteristics of the game in the video Stream that are exploited to interpret soccer. For example, they correspond to how players are distributed in the image or the part of the field that is filmed. Finally, we show how these interpretative semantic features can be used to set up and train a semantic-based decision tree classifier for major game events with a restricted amount of training data. The main advantages of our semantic approach are that it only requires the video feed of the main Camera to extract the semantic features, with no need for Camera calibration, field homography, player tracking, or ball position estimation. While the automatic interpretation of sports games remains challenging, our approach allows us to achieve promising results for the semantic feature extraction and for the classification between major soccer game events such as attack, goal or goal opportunity, defense, and middle game.

Shaoyi Chien - One of the best experts on this subject based on the ideXlab platform.

  • ressp a 5 877 tops w reconfigurable smart Camera Stream processor
    Custom Integrated Circuits Conference, 2011
    Co-Authors: Weikai Chan, Yuhsiang Tseng, Peikuei Tsung, Tzuder Chuang, Yimin Tsai, Weiyin Chen, Lianggee Chen, Shaoyi Chien
    Abstract:

    A 5.877 TOPS/W Reconfigurable Smart-Camera Stream Processor is implemented in 90nm CMOS technology. A reconfigurable hardware architecture with heterogeneous Stream processing and subword-level parallelism is implemented to accelerate the vision processing for smart-Camera applications. The area efficiency reaches 111.329 GOPS/mm2. The power efficiency and area efficiency are 4.5x to 33.0x and 3.8x to 74.2x better than the state-of-the-art works, respectively.

  • CICC - ReSSP: A 5.877 TOPS/W Reconfigurable Smart-Camera Stream Processor
    2011 IEEE Custom Integrated Circuits Conference (CICC), 2011
    Co-Authors: Weikai Chan, Yuhsiang Tseng, Peikuei Tsung, Tzuder Chuang, Yimin Tsai, Weiyin Chen, Lianggee Chen, Shaoyi Chien
    Abstract:

    A 5.877 TOPS/W Reconfigurable Smart-Camera Stream Processor is implemented in 90nm CMOS technology. A reconfigurable hardware architecture with heterogeneous Stream processing and subword-level parallelism is implemented to accelerate the vision processing for smart-Camera applications. The area efficiency reaches 111.329 GOPS/mm2. The power efficiency and area efficiency are 4.5x to 33.0x and 3.8x to 74.2x better than the state-of-the-art works, respectively.

Matthew Turk - One of the best experts on this subject based on the ideXlab platform.

  • TranslatAR: A mobile augmented reality translator
    2011 IEEE Workshop on Applications of Computer Vision WACV 2011, 2011
    Co-Authors: Victor Fragoso, Jim Kleban, Shane Zamora, Steffen Gauglitz, Matthew Turk
    Abstract:

    We present a mobile augmented reality (AR) translation system, using a smartphone's Camera and touchscreen, that requires the user to simply tap on the word of interest once in order to produce a translation, presented as an AR overlay. The translation seamlessly replaces the original text in the live Camera Stream, matching background and foreground colors estimated from the source images. For this purpose, we developed an efficient algorithm for accurately detecting the location and orientation of the text in a live Camera Stream that is robust to perspective distortion, and we combine it with OCR and a text-to-text translation engine. Our experimental results, using the ICDAR 2003 dataset and our own set of video sequences, quantify the accuracy of our detection and analyze the sources of failure among the system's components. With the OCR and translation running in a background thread, the system runs at 26 fps on a current generation smartphone (Nokia N900) and offers a particularly easy-to-use and simple method for translation, especially in situations in which typing or correct pronunciation (for systems with speech input) is cumbersome or impossible.

  • WACV - TranslatAR: A mobile augmented reality translator
    2011 IEEE Workshop on Applications of Computer Vision (WACV), 2011
    Co-Authors: Victor Fragoso, Jim Kleban, Shane Zamora, Steffen Gauglitz, Matthew Turk
    Abstract:

    We present a mobile augmented reality (AR) translation system, using a smartphone's Camera and touchscreen, that requires the user to simply tap on the word of interest once in order to produce a translation, presented as an AR overlay. The translation seamlessly replaces the original text in the live Camera Stream, matching background and foreground colors estimated from the source images. For this purpose, we developed an efficient algorithm for accurately detecting the location and orientation of the text in a live Camera Stream that is robust to perspective distortion, and we combine it with OCR and a text-to-text translation engine. Our experimental results, using the ICDAR 2003 dataset and our own set of video sequences, quantify the accuracy of our detection and analyze the sources of failure among the system's components. With the OCR and translation running in a background thread, the system runs at 26 fps on a current generation smartphone (Nokia N900) and offers a particularly easy-to-use and simple method for translation, especially in situations in which typing or correct pronunciation (for systems with speech input) is cumbersome or impossible.

Xiuwen Yin - One of the best experts on this subject based on the ideXlab platform.

  • Towards cost-efficient cloud resource management for large scale Camera Stream analysis
    Alexandria Engineering Journal, 2020
    Co-Authors: Weiwen Zhang, Jianqi Liu, Xiaochun Cheng, Wangkit Wong, Xiuwen Yin
    Abstract:

    Abstract A large number of public Cameras are connected to the Internet, providing valuable information to modern society. We consider a cloud-based system that analyzes the Streaming data from the Cameras. It consists of a manager to dispatch jobs and workers to process the jobs. The workers are bound with various instances from public cloud that have different performance in terms of average frame rate and monetary cost. In this paper, we investigate the resource management on how to dispatch jobs to workers. First, we aim to maximize the minimum average frame rate of workers while satisfying the budget of renting computing instances from public cloud. Second, we aim to minimize the cost while satisfying the desired frame rate for Stream analysis. For the former, we transform the maximization problem into a knapsack problem; for the latter, we transform the minimization problem as a bin packing problem. We develop efficient greedy algorithms to solve optimization problems with theoretical bounds. Simulation results indicate the tradeoff between performance and cost: as the budget of renting cloud instances, the benefit diminishes in achieving high average frame rate; as the number of Cameras increases, more additional cost will incur to provide the desired frame rate.

Mattias Wahde - One of the best experts on this subject based on the ideXlab platform.

  • ICAART (Revised Selected Papers) - A Method for Binarization of Document Images from a Live Camera Stream
    Lecture Notes in Computer Science, 2015
    Co-Authors: Mattias Wahde
    Abstract:

    This paper describes a method for binarization of document images from a live Camera Stream. The method is based on histogram matching over partial images referred to as tiles. A method developed previously has been applied successfully to images with artificially added noise. Here, an improved method is presented, in which the user has more direct control over the specification of the binarizer. The resulting system is then taken a step further, by considering the more difficult case of binarization of live Camera images. It is demonstrated that the improved method works well for this case, even when the image Stream is obtained using a slightly modified low-cost web Camera with low resolution. For typical images obtained this way, a standard OCR reader is capable of reading the binarized images, detecting around 87.5i¾?% of all words without any error, and with mostly minor, correctable errors for the remaining words.

  • a method for binarization of document images from a live Camera Stream
    International Conference on Agents and Artificial Intelligence, 2014
    Co-Authors: Mattias Wahde
    Abstract:

    This paper describes a method for binarization of document images from a live Camera Stream. The method is based on histogram matching over partial images referred to as tiles. A method developed previously has been applied successfully to images with artificially added noise. Here, an improved method is presented, in which the user has more direct control over the specification of the binarizer. The resulting system is then taken a step further, by considering the more difficult case of binarization of live Camera images. It is demonstrated that the improved method works well for this case, even when the image Stream is obtained using a slightly modified low-cost web Camera with low resolution. For typical images obtained this way, a standard OCR reader is capable of reading the binarized images, detecting around 87.5i¾?% of all words without any error, and with mostly minor, correctable errors for the remaining words.