close menu

Applied ML seminars – Anomaly detection (אירוע)

אני שמח להזמין אתכם לאירוע השני שלנו בסדרת Applied ML seminars שנעשים בשיתוף עם Applied Materials (לינק לאירוע הראשון). בכל אירוע מסוג זה, ניקח נושא אחד שמעניין את הקהילה ונדבר עליו בהרחבה מכמה זוויות שונות. לאירוע הראשון בחרנו לדבר על נושא שלא מעט מחברי הקהילה עוסקים בו: Anomaly detection. באירוע נדבר על גישות שונות למודלים בתחום ונשמע הרצאות משני דוברים בנושא.

האירוע יתקיים ב-16.2 בשעה 17:00

ניתן להירשם לאירוע דרך הלינק הזה (הוסיפו ליומן).

הרצאה ראשונה:

 

Speaker: Jonathan Laserson, Lead AI Strategist at Zebra Medical Vision

Title: "Augmentation is all you need".

Abstract:
Data augmentation has been around since the early days of deep learning. This lecture
will tell its story, how it grew from a trivial trick to increase the size of our training set, to
a key ingredient in the quest for unsupervised learning, one of the few elements of deep
learning that still need a human touch. I'll discuss the role of augmentation in state of the
art methods like anomaly detection and contrastive learning. I'll demonstrate how the
choices we make regarding the type of augmentation we allow during training make a
profound statement about our data.

הרצאה שניה:

Speaker: Ran Yacoby, Algorithm Developer at Applied Materials

Title: “Self-supervised learning and anomaly detection in the semiconductor industry.”

Abstract:

“The fabrication of integrated circuits (ICs) in the semiconductor industry is a highly complex process, involving hundreds of steps that gradually grow patterned nano-scale structures on top of silicon wafers. Each of these steps is prone to a variety of defects that can reduce the quality of the IC or ruin it entirely. Properly detecting and classifying such defects is crucial for maintaining the yield of the manufacturing process. The classical approaches for addressing these problems are in the process of being replaced by more powerful deep-learning algorithms. In this talk, I will focus on self-supervised representation learning techniques in the context of unsupervised anomaly detection. Specifically, representations learned through contrastive-learning recently achieved remarkable improvements in semi-supervised classification and object detection tasks, and we will discuss their application to the anomaly detection use-case.”

עוד בנושא:
افلام سكس تبادل secretsporn.com شيميل سحاق sexy mms in hindi partyporntrends.com xvideos indian girlfriend コスプレ一本勝負第九十九試合! 『正統派美少女連続絶頂生ハメ種付け舌出しアヘ顔セックス』 javidol.org アナル責め 動画 sexy video bp hindi fuckvidstube.com xxxvideosdownload boobs sucking hot pornxvideos.info jat sex
ref tube tubeporncity.info sexy brest pics xx videos hd dudano.mobi indian secret sex vedios delhi video sexy tubedessert.mobi nauty american sex www kamapisachi videos rajwap.biz sex blue photo gj-bu hentai hentaifreeporno.com furrie hentai
mallu mom sex fullindiantube.com ladies tamil جنس غريب pornowap.mobi اجمل قصص الجنس teen.com joysporn.mobi ilayaraja wife stars-056 javunsensored.net 下着モデルをさせられて… 奥田咲 xxx com bengali ruperttube.net yespronplease