NIPS workshop: Machine Learning for Next Generation Computer Vision Challenges

Call for contributions for a NIPS workshop: Machine Learning for Next Generation Computer Vision Challenges



Submission deadline: 23:59, GMT, 18th October, 2010

Workshop overview:


This workshop seeks to excite and inform researchers to tackle the next level of problems in the area of Computer Vision. The idea is to both give Computer Vision researchers access to the latest Machine Learning research, and also to communicate to researchers in the machine learning community some of the latest challenges in computer vision, in order to stimulate the emergence of the next generation of learning techniques. The workshop itself is motivated from several different points of view:

1. There is a great interest in and take-up of machine learning techniques in the computer vision community. In top vision conferences such as CVPR, machine learning is prevalent: there is widespread use of Bayesian Techniques, Kernel Methods, Structured Prediction, Deep Learning, etc.; and many vision conferences have featured invited speakers from the machine learning community.
2. Despite the quality of this research and the significant adoption of machine learning techniques, often such techniques are used as “black box” parts of a pipeline, performing traditional tasks such as classification or feature selection, rather than fundamentally taking a learning approach to solving some of the unique problems arising in real-world vision applications.
3. Beyond object recognition and robot navigation, many interesting problems in computer vision are less well known. These include more complex tasks such as joint geometric/semantic scene parsing, object discovery, modeling of visual attributes, image aesthetics, etc.
4. Even within the domain of “classic” recognition systems, we also face significant challenges in scaling up machine learning techniques to millions of images and thousands of categories (consider for example the ImageNet data set).
5. Images often come with extra multi-modal information (social network graphs, user preference, implicit feedback indicators, etc) and this information is often poorly used, or integrated in an ad-hoc fashion.

This workshop therefore seeks to bring together machine learning and computer vision researchers to discuss these challenges, show current progress, highlight open questions and stimulate promising future research.

Call for Papers


Papers are sought in the following areas:

* Use of multi-modal information in image tasks (e.g., text, GPS tags, timestamps, social network, implicit feedback, audio, user preferences)

* Image tasks beyond object classification — that is, novel applications (comprehensive scene understanding, object discovery, attribute learning, aesthetic analysis, modeling of the collective structure of large-scale image datasets, etc.)

* Novel learning techniques and features especially suited for the above applications

* Papers that emphasize on integrated learning approaches, in contrast to solving any issues purely via complex software engineering (i.e., by chaining standard methods).

* Methods that are truly scalable to millions of images and/or to large video repositories, which now dominate many vision tasks.

* Algorithms that really push the boundaries of Machine Learning for Computer Vision tasks, or applications which really push the boundaries of both disciplines are particularly sought.

The program committee will review papers and provide suggestions for either a poster or oral presentation. Note that scientific contribution is a must; however, we encourage preliminary approaches that partially solve a challenging issue, or solutions that target a problem of interest but are not necessarily state-of-the-art in terms of performance (e.g., a method that scales to 1 trillion images on a mobile phone, but is 2% behind the winner on the latest vision challenge so would not necessarily be considered ‘state of the art’). The aim of the workshop is to look to the future, as much as it is to demonstrate successes of the (recent) past.

Call for Demos/Projects


We especially solicit posters and/or demos from projects (e.g. internal, NSF funded, EU projects). This can be from projects near completion — an opportunity to show the community what challenges were addressed and demonstrate and software / datasets / systems that were produced. Alternatively, these can form outlines, ideas, open problems. The idea is to raise awareness of all activities in the joint area of machine learning / computer vision among as many researchers as possible. We will aim to accommodate as many relevant demos/project posters as possible.

Our overall aim: is to promote fruitful discussion among researchers from both communities, to raise awareness of work / challenges / projects / datasets, and to provide a relaxed environment in which to discuss these aspects. We are not aiming at a processional mini-conference, the outcome of the workshop should be more than a list of papers to go and read: hopefully you will have new contacts and new research ideas to get very excited about.




Submission deadline: 23:59, GMT, 18th October, 2010