<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=UTF-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p><br>
    </p>
    <p>Hi everyone - <br>
    </p>
    <p>Please come to our Colloquium today with our speaker Ming-Hsuan
      Wang:</p>
    <p>
      <blockquote type="cite">
        <p class="MsoNormal" style="text-align:justify;margin:0in 0in
          0.0001pt;font-size:12pt;font-family:Calibri,sans-serif"><i><span
              style="font-family:Arial,sans-serif">Host: Tobias Höllerer
              & William Wang</span></i></p>
        <p class="MsoNormal" style="text-align:justify;margin:0in 0in
          0.0001pt;font-size:12pt;font-family:Calibri,sans-serif"><b><span
              style="font-family:Arial,sans-serif"> </span></b></p>
        <p class="MsoNormal" style="text-align:justify;margin:0in 0in
          0.0001pt;font-size:12pt;font-family:Calibri,sans-serif"><b><span
              style="font-family:Arial,sans-serif">Title: Learning to
              Track and Segment Objects in Videos</span></b><b><span
              style="font-family:Arial,sans-serif;color:black"> </span></b></p>
        <p class="MsoNormal" style="text-align:justify;margin:0in 0in
          0.0001pt;font-size:12pt;font-family:Calibri,sans-serif"><span
            style="font-family:Arial,sans-serif"> </span></p>
        <p class="MsoNormal" style="text-align:justify;margin:0in 0in
          0.0001pt;font-size:12pt;font-family:Calibri,sans-serif"><b><span
              style="font-family:Arial,sans-serif">Abstract: </span></b></p>
        <span style="font-family:Arial,sans-serif">Visual tracking and
          object segmentation are two important tasks for understanding
          the motion contents in videos that find numerous applications.
          The typical setup for visual tracking is to initialize the
          target of interest with a bounding box in the first frame and
          the goal is determine locations of the target throughout the
          frames. While significant progress has been made in the past
          decades, object tracking remains a difficult problem due to
          large appearance change, shape deformation, light variation,
          and occlusions. In the first part, I will present our recent
          findings on tracking single and multiple objects in videos. I
          will also discuss numerous applications based on visual
          tracking. On the other hand, a closely related problem to how
          to segment objects in videos. </span><span
          style="font-family:Arial,sans-serif;color:black">Online video
          object segmentation is a difficult task as it entails to
          process the image sequence timely and accurately. To segment a
          target object through the video, numerous methods based on
          deep learning have been developed by heavily fine-tuning on
          the object mask in the first frame, which is time-consuming
          for online applications.  In the second part, I will present a
          fast and accurate video object segmentation algorithm with
          applications. </span><span
          style="font-family:Arial,sans-serif">When time allows, I will
          briefly discuss our work on low-level and high-level vision
          tasks including frame interpolation, image synthesis, and
          style transfer.  </span></blockquote>
      I will bring Ming-Hsuan by the HFH lab around 5pm for a quick tour
      and chat - would be great if you could briefly show your work! <br>
    </p>
    <p>Also, please all go to the student lunch tomorrow (Thu) at noon:
      <br>
    </p>
    <p><a class="moz-txt-link-freetext" href="https://docs.google.com/spreadsheets/d/1mptIBhL8hvEYK8OeFBrBGbahSgym1pvQW-Y8alqgjZI/edit#gid=1873430703">https://docs.google.com/spreadsheets/d/1mptIBhL8hvEYK8OeFBrBGbahSgym1pvQW-Y8alqgjZI/edit#gid=1873430703</a><br>
    </p>
    <p>Cheers,</p>
    <p>    Tobias</p>
    <p><br>
    </p>
    <pre class="moz-signature" cols="72">-- 
Tobias Hollerer
Professor, Department of Computer Science
University of California, Santa Barbara, CA 93106-5110

<a class="moz-txt-link-abbreviated" href="mailto:holl@cs.ucsb.edu">holl@cs.ucsb.edu</a>, Office: (805)284-9395, Fax: (805)893-8553 </pre>
  </body>
</html>