[4th July, 11:30 ] Dr. Ramazan S. Aygun: Unsupervised Speaker Identification for TV News

Place: TH:A-1455

Television (TV) networks produce a tremendous amount of information every day. Identifying the speakers throughout a video would help to analyze and understand the video content. Previous research has usually identified speakers on pre-trained faces of famous people for TV shows and movies.
News videos are challenging because new faces (or people) often appear. By using an unsupervised method, this paper proposes to label speakers using just the available information in the news video without external information. Our proposed framework segments the audio by speaker, parses closed captions for speaker names, identifies talking persons, and performs optical character recognition for speaker names.
The presentation will show

  • how speaker diarizarion, face recognition, face landmarking, natural language processing, and optical character recognition tools can be effectively used utilized for speaker identification,
  • how speakers who are not famous could be recognized using different modalities, and
  • present results for identifying speakers for CNN news with overall accuracy of 63.6% including speakers just appearing once.

About the speaker

Read more: [4th July, 11:30 ] Dr. Ramazan S. Aygun: Unsupervised Speaker Identification for TV News

[27th June, 14:00] Bradford Cross: Machine Learning Startups

Place: TH:A-1455

Brandford will tell us something about his experience (not only) with machine learning startups.

Bradford is a founding partner at DCVC, the world's leading machine learning and big data venture capital fund, where he starts and invests in machine learning ventures. His current focus is fintech and applications of computational bio, and machine vision.

He has founded two machine learning startups since 2009. At Prismatic, he and his team used Machine learning for personalized ranking based on social and content interaction, and Natural Language Processing for topic and entity classification. At Flightcaster, they used machine learning for predicting the state of the real time global air traffic network using FAA, carrier, and weather data.

He has been a hedge fund investor since 2002, starting with statistical value and momentum strategies at O’Higgins Asset Management, and a venture investor since 2010 as a founding partner of Data Collective.

He has spent 7 years building statistical trading strategies, 6 years building machine learning startups, and 8 years on systems engineering including 2 years working on distributed systems at Google. He studied Computer Engineering and Finance at Virginia Tech, and Mathematics at Berkeley.

Summer Camp 2016

No summer plans yet?

Keen to work on interesting projects in areas of Artificial Intelligence, Machine Learning, Data Mining and Big Data?

Well, this is the right place to be!

DataLab introduce Summer Camp 2016 for students interested in Artificial Intelligence, Data Mining, Machine Learning and Big Data.

Pre-register here.

Read more: Summer Camp 2016

[May 5, 13:30] Pavel Brazdil: Meta-learning and text mining

Datalab is happy to invite you to meet prof. Pavel Brazil, researcher in the field of meta learning and data mining. Pavel established strong research group in the Laboratory of AI and Decision Support in Porto, Portugal.

Place: T9:364 (New Building of CTU)

Pavel will talk about meta learning, text summarization and sentiment analysis. There will be 20 minutes discussion.

About the Speaker

Read more: [May 5, 13:30] Pavel Brazdil: Meta-learning and text mining

Datalab Summer Camp presentations (batch 2)

Datalab invites you to the second batch of Summer Camp final presentations. Presentations will start on Thursday, Nov 26th, 2pm, room 364, Thakurova 9, Prague and finish approximately at 15:30.

Here are the projects to be presented:

  • Jan Lukány, Log Mining (Lely)
  • Alena Moravová, Clockwork Recurrent Neural Networks (GoodAI)
  • Jakub Švehla, Data Clustering (Lely)
  • Jiří Nádvorník, Universal Artificial Intelligence AIXI (GoodAI
  • Petr Prokop, Analysis of settings of automated milking system (Lely)
  • Sergii Stamenov, Doc2vec and article popularity prediction (Predictable.ly)
  • Josef Dvořák, Improving Matchmaking in CWI: Cold Start Problem Mitigation in Recommender Systems (research project funded by FIT CTU

Datalab invites you to the second batch of Summer Camp final presentations. Presentations will start on Thursday, Nov 26th, 2pm, room 364, Thakurova 9, Prague and finish approximately at 15:20.

Follow Us

Copyright (c) Data Science Laboratory @ FIT CTU 2014–2016. All rights reserved.