当前位置:首页>>博士后之家>>国外博士后招聘>>正文内容

英国伦敦大学学院视觉数据分析中的机器学习方向博士后职位招聘

2018年07月17日
来源:知识人网
摘要:Research Assistant in Machine Learning for Visual Data Analysis University College London - UCL Electronic & Electrical Engineering

  Research Assistant in Machine Learning for Visual Data Analysis

  University College London - UCL Electronic & Electrical Engineering

博士后申请

  As part of our work within a number of EPSRC and EU-funded projects, the UCL Electronic & Electrical Engineering , Communications and Informations Systems Group invites applications for a Research Assistant in Machine Learning for Visual Data Analysis.

  For the last 50 years, the holy grail of machine learning with visual data has been to translate pixels to concepts, e.g., classify a pixel-domain video according to its contents ('tennis match', 'cooking show', 'person driving a van'...) or find video scenes that are semantically similar to the contents of a given query video. However, pixel-domain video representations are in fact known to be cumbersome for machine learning, due to : limited frame rate, too much redundancy between successive frames, calibration problems under irregular camera motion, blurriness due to shutter adjustment under varying illumination and very high power requirements. Inspired by biological vision, new input modalities are now beginning to be considered for visual data analysis, e.g., neuromorphic visual sensors (a.k.a., silicon retinas), or compressed-domain motion and RGB information from video codecs like MPEG/ITU-T AVC/H.264 and HEVC instead of uncompressed (pixel-domain) video. At the same time, exciting developments in transfer learning and discriminative domain adaptation allow for knowledge transfer from one data modality to another, thereby opening new opportunities to advance the state-of-the-art in resource-efficient visual data analysis that can be deployed in practical systems.

  We are looking for a talented research assistant to join our team and help us fulfil the projects' goals, producing quality research in transfer learning or discriminative domain adaptation for visual data analysis and recognition problems, including but not limited to, the problems and data modalities mentioned above. The work will involve design, development and implementation work and publishing high quality research papers in high-ranked conferences and journals. The successful candidate will work within an established research team in the Communications and Information Systems Group, led by Dr Yiannis Andreopoulos. The position is available from September 2018 for 24 months in the first instance.

  Applicants are required to have a Masters degree (or 4 or 5-year undergraduate degree) in Computer Science, Electronic Engineering or a related field. Fluency in Python and Matlab programming evidenced by previous usage in research papers is essential, as is an understanding of data science and machine learning, evidenced by high marks in related graduate -level modules or completion of related online courses in Coursera or similar. It is desirable that applicants have some exposure in the use of machine learning libraries like Caffe, Tensorflow, Keras or similar, evidenced by extensive use in data problems, competitions or research publications. The full person specification can be found in the job description.

  UCL vacancy reference: 1732339

  Applicants should apply online. To access further details about the position and how to apply please click on the ‘Apply’ button below.

  Interested applicants are encouraged to make informal enquiries about the post to Dr Yiannis Andreopoulos (i.andreopoulos@ucl.ac.uk). Any questions regarding the application process should be directed to Vicky Coombes (v.coombes@ucl.ac.uk)

  Latest time for the submission of applications: 23:59

  Interview Date: TBC

  UCL Taking Action for Equality

  Our department is working towards an Athena SWAN award. We are committed to advancing gender equality within our department.

  以上内容由知识人网博士后申请咨询老师整理发布,希望能给遇到此类状况的申请者们一个清晰的方向。想了解更多关于国外博士后博士后招聘的相关资讯,可以登录我们的官方网站:www.zsr.cc。