ICLR
ICLR, full name of the International Conference on Learning Representations, the Chinese translation for the "International Conference on Characterization of learning." The conference was founded by Yoshua Bengio and Yann LeCun led, in 2013 held its first meeting.
ICLR founder purposes:
As we all know, the application performance characterization data for machine learning have an important impact. Characterization of the rapid development of learning also accompanied by many problems, such as how we can be more effective characterization of the meaning and better learn from the data. We started to explore this area, including in-depth study, characterization study, a measure of learning, kernel learning, combined model, very linear structure prediction and non-convex optimization problem. Despite the characterization of learning for vision, speech, audio, and plays a vital role in the field of NLP and Machine learning including, the current lack of a place that allows the exchange of scholars in the field to share topics of interest. ICLR aim is to fill this gap.
ICLR Highlights:
Open Review evaluation system 1 , in accordance with the provisions of:
All papers will disclose the names and other information submitted, and accepted all the peer reviews and questions (open peer review), or any scholar can evaluate the real name or anonymously papers. And at the end of the public review, the authors can be adjusted and modified to paper.
ICLR official submission entrance: OpenReview.net
OpenReview.net public review system is a University of Massachusetts Amherst College Andrew McCallum is ICLR 2013 led by founder, uphold public peer-reviewed, published, open source, open discussion, open the boot, publicly recommended, such as the eight open API and open source in principle, the support of institutions of Facebook, Google, NSF and the University of Massachusetts Amherst center.
CVPR
CVPR, the full name of IEEE Conference on Computer Vision and Pattern Recognition , Chinese translation of "IEEE International Conference on Computer Vision and Pattern Recognition." In 1983, the first session CVPR conference in Washington by the Kanade and Dana Ballard held thereafter every year in the United States held territory.
CVPR) is the IEEE's annual academic conference, the main content of the meeting is computer vision and pattern recognition technology. Generally held in June, location is often in the United States western, central or eastern regions. For example, 2013 was held in Portland, be held in Colombia in 2014, Boston in 2015, 2016 in Las Vegas, 2017 in Hawaii.
CVPR, tied together with ICCV and ECCV computer vision will be the top three . In recent years, every year some 1,500 participants, the number of papers included about 300 general, the overall rate of collection of <30%, the proportion of papers oral report of <5% . CVPR generally use "double blind review" usually require re-reading a paper by three reviewers, the final decision by the area chair the meeting was received. The meeting will have a fixed annual discussion topic. Companies can also obtain the opportunity to show at the venue by sponsoring the meeting.
ECCV
ECCV, full name of the European Conference on Computer Vision, Chinese translation of "European Conference on Computer Vision International." Two years to convene a. The number of papers a year to hire about 300, about 27 percent acceptance rate in 2010, hired a source of dissertation for the United States, Europe and other leading laboratories and research institutes, mainland China almost 10-20 papers are hired each year.
ICML
ICML, full name of the International Conference on Machine Learning, Chinese translation of the "International Conference of Machine Learning", organized by the International Machine Learning Society (IMLS), once a year. NIPS tied with machine learning will be the top two.
NIPS
NIPS, full name of the Conference and Workshop on Neural Information Processing Systems, Chinese translation of "Neural Information Processing System" of the General Assembly, an international conference on machine learning and computational neuroscience.
The General Assembly is held annually in December, organizers NIPS Foundation. After the meeting usually in Proceedings of the 2 early version of the meeting, the name of the conference "Advances in Neural Information Processing Systems."
NIPS standard machine learning sessions (such as ICML) different, a considerable part of NIPS conference on neuroscience, but because the content is still the main machine learning, all is still a lot of people regarded as the best conference of machine learning One.
References:
[. 1] https://zhuanlan.zhihu.com/p/51749414
[2] https://www.leiphone.com/news/201704/CdCLonir2okijXtg.html
[. 3] HTTPS: //blog.csdn. net / program_developer / article / details / 72846359
The current review paper is divided into single-blind (single-blind review), double-blind (double-blind review) and open review (open review) and other forms. Meaning ① very simple single-blind review, that review to know the author's name, school and other identity information, but does not know who the author of the review paper; ② and double blind review process is neither know each other's identity. The two main ways to approach many academic conferences and journals reviewed papers. The advantages of single-blind review is very obvious that the assessment in anonymity, lets review from stress, but knowing the author reviewed the information paper, it is very likely to produce stereotypes, not enough to produce an objective assessment results. For example, for the more famous scholar, quality review papers will produce presets. The double blind review process is to significantly reduce the effect of the additional personal information brings, however, the assessment is very likely to determine the author from the wording, themes. ↩