<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=Windows-1252">
</head>
<body>
<div dir="ltr">
<div></div>
<div>
<div dir="ltr"><br>
</div>
</div>
<div id="id-1309032a-ecc7-4ebd-8b06-4abe7ced2a78" class="ms-outlook-mobile-reference-message">
<div dir="ltr">
<div dir="ltr">
<table align="center" bgcolor="#fff" border="0" cellpadding="0" cellspacing="0" width="600" style="font-family:Arial,sans-serif">
<tbody>
<tr>
<td><img src="https://www.egr.uh.edu/sites/www.egr.uh.edu/files/enews/2022/images/dissertation1.png" alt="Dissertation Defense Announcement at the Cullen College of Engineering" width="600" height="171">
<table align="center" bgcolor="#ffffff" border="0" cellpadding="10" cellspacing="0">
<tbody>
<tr>
<td align="center" style="padding:40px 20px 10px">
<div style="font-size:24px; color:rgb(200,16,46); line-height:28px"><strong>Better Generalization with Less Human Annotation<br>
</strong></div>
<div style="margin-top:5px; line-height:22px"><strong>USING META-LEARNING AND SELF-SUPERVISED LEARNING FOR IMAGE ANALYSIS</strong></div>
<div style="margin:30px 0px; line-height:20px">
<div style="font-size:18px; margin-bottom:5px"><b>Pengyu Yuan</b></div>
<div style="font-size:14px; line-height:20px">
<p style="font-family:Arial,Helvetica,sans-serif; line-height:22px; margin:0px 0px 5px">
July 29, 2022; 12:00 PM - 14:00 PM (CST)<br>
Location: Online on Teams<br>
Teams: <a href="https://urldefense.com/v3/__https://teams.microsoft.com/l/meetup-join/19*3ameeting_NTM4MzcxNjEtNzNhYi00ZmI3LWExY2YtOGRhY2YwZGEzNWFm*40thread.v2/0?context=*7b*22Tid*22*3a*22170bbabd-a2f0-4c90-ad4b-0e8f0f0c4259*22*2c*22Oid*22*3a*2283979ffb-3903-4443-94b4-7f9efcbda9a3*22*7d__;JSUlJSUlJSUlJSUlJSUl!!LkSTlj0I!CMn4EUIvMRFMxVYUA1_ZlPshV2SJeotKMadq8Jnn1gMdNdPZPvH1w8HbuaYjItg1pny9s2QMFHrWPT3wFgzxHg4K$" target="_blank" rel="noreferrer" style="color:rgb(200,16,46)">https://teams.microsoft.com/l/meetup-join/...</a></p>
</div>
</div>
<div style="font-size:14px; line-height:20px">
<p style="font-family:Arial,Helvetica,sans-serif; line-height:22px; margin:0px 0px 5px">
<strong>Committee Chair:</strong><br>
Hien Van Nguyen, Ph.D.<br>
</p>
</div>
<div style="font-size:14px; line-height:20px">
<p style="font-family:Arial,Helvetica,sans-serif; line-height:22px; margin:0px 0px 20px">
<strong>Committee Members:</strong><br>
Jiefu Chen, Ph.D. | Zhu Han, Ph.D. | Xuqing Wu, Ph.D. | Stephen T. Wong, Ph.D.</p>
</div>
</td>
</tr>
<tr>
<td style="padding:0px 20px 20px">
<p style="font-family:Arial,Helvetica,sans-serif; font-size:16px; line-height:22px; margin:15px 0px; color:rgb(200,16,46)">
<strong>Abstract</strong></p>
<p style="font-family:Arial,Helvetica,sans-serif; font-size:14px; line-height:22px; margin:15px 0px">
<span style="font-family:Arial,sans-serif; font-size:10.5pt">Deep neural networks require a large amount of annotated training data to generalize well. Unfortunately, such large training data is difficult to obtain in medical or geophysical domains due to many
reasons, including the high annotation cost, privacy concerns, and physical constraints. To address the data efficiency issue in training the deep learning model, we proposed the solutions in two directions: meta-learning and self-supervised learning. Meta-learning
tries to generate a robust model that can learn to quickly adapt to new tasks with minimal labeled samples. It is also called “learning to learn”, which acquires fast adaptation capability over a collection of related tasks and uses it to improve its future
learning performance. Self-supervised learning, on the other hand, tries to leverage all useful information from the unlabeled training data itself. It can improve feature representation by solving a pretext task, or directly solving the unsupervised target
task which shares the same task structure as the self-supervised task. In the first part of this work, we introduce the meta-learning setting, develop a meta-learning framework AGILE+ </span><span style="font-family:Arial,sans-serif; font-size:10.5pt">to deliver
the efficient rat brain cell classifier, and study the first arrival picking problem and solve the domain shift problem with less human interaction. In the second part, we discuss different self-supervised learning settings, introduce the 3D self-supervised
image patch reconstruction task to significantly improve the incidental lung nodule classification accuracy for the data-hungry 3D Vision Transformer (3D-ViT) model, and propose the self-supervised learning model Blind-Trace Network (BTN) for the application
in the unsupervised seismic interpolation task. </span></p>
</td>
</tr>
</tbody>
</table>
</td>
</tr>
<tr>
<td><img src="https://www.egr.uh.edu/sites/www.egr.uh.edu/files/enews/2022/images/dissertation2.png" alt="Engineered For What's Next" width="600" height="82"></td>
</tr>
</tbody>
</table>
</div>
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex; border-left:1px solid rgb(204,204,204); padding-left:1ex">
<div>
<div dir="ltr">
<div id="gmail-m_-7801825294218122341id-f8014617-5efb-480f-ab68-f3d9cb90bc3f">
<div dir="ltr">
<div>
<div dir="ltr">
<div dir="ltr">
<div>
<div dir="ltr">
<div dir="ltr"></div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
</div>
</div>
</div>
</div>
</body>
</html>