cool hit counter AIChallenger Global AI Challenge concludes with winning teams sharing $2 million in prizes_Intefrankly

AIChallenger Global AI Challenge concludes with winning teams sharing $2 million in prizes


【TechWeb报道】12月21日消息,由创新工场、搜狗、今日头条主办,历时约130天,吸引65国家、上万人报名参加的首届 “AI Challenger全球AI挑战赛”今天落下帷幕。AI Challenger视觉和翻译两大类共五个赛道,各自决出了最终的冠军,以及亚军、季军、优秀奖。获奖团队合计分享了超过200万人民币的奖金。

据了解,“AI Challenger全球AI挑战赛”由中关村管委会指导,创新工场、搜狗、今日头条联合创建,旨在从科研角度出发,满足学术界对高质量数据集的需求,推进人工智能在科研与商业领域的结合,促进世界范围内人工智能研发人员共同探索前沿领域的技术突破及应用创新。

The competition was officially launched to the world on August 14 this year, and is divided into five tracks: human bone key point detection competition, image Chinese description competition, scene classification competition, English-Chinese machine text translation competition and English-Chinese machine simultaneous interpretation competition. The competition is based on a joint AB double set test. Open dataset A for testing first, followed by dataset B for testing. Weekly winners are awarded every fortnight and the teams with outstanding results in each competition are ranked according to the final results list to make the finalists.

After about 130 days, the competition saw the final moment of the defense, with more than 20 teams presenting and defending for 25 minutes each, and the winning teams in all five tracks were decided. The competition attracted over a million people from 120 countries worldwide, with tens of thousands of participants from 65 countries forming a total of 8,892 teams. A total of 347 colleges and universities and about 273 companies were covered.

It is understood that the winners of the competition, in addition to the generous prizes, will also have the opportunity to gain access to jobs, internships or investments with the three organizers, as well as share their winning experiences at top international academic conferences.

三家主办方认为,在人工智能领域,数据的质和量是科学研究与产品技术研发的核心。高质量训练数据对机器学习模型的建立和优化有关键性的作用。建立大规模、高水准的标注数据集,是推动AI科研和技术前进的驱动力。

According to the introduction, the competition opened up a million-scale computer vision dataset and a 10-million-scale machine translation dataset, including: more than 10 million Chinese and English translation data, 700,000 human skeleton key point annotation data, 300,000 picture scene annotation and semantic description data. This is the largest publicly available research dataset in the country to date.

创新工场董事长兼CEO李开复表示,希望未来三年,AI Challenger能消除中美年轻顶尖人才之间的差距,从我们这里走出去的优秀选手,最终成长成中国乃至世界AI产业的领军人物。

搜狗CEO王小川表示,很欣喜地看到众多技术牛人参与了此次大赛并取得了优秀的成绩。不仅对自然语言处理提供了新的技术思路,也给了我们很多启发。未来希望每年都能通过大赛聚集优秀人才,在AI领域取得更多突破,推动中国人工智能更快更好地发展 。

Today's headline founder Zhang Yiming said, AI Challenger for the world's top research talent in the field of artificial intelligence, we provide the most powerful data support for all participants, so that artificial intelligence can better serve the creation and communication of information.

Wang Yonggang, executive committee member of the competition and vice president of Innovation Works AI Engineering Institute, put forward a more specific annual goal, hoping that in 2018, AI Challenger will attract more international and domestic high-level participating teams and establish a sharing and exchange platform for algorithms, code, papers, data and competitions.

据了解,未来三年,主办方将投入数千万基金,解决数据集缺失的问题,为人工智能科研提供海量数据及算法竞赛、人才交流平台,辅以学术界和产业界专家指导,建设和发布更大规模的AI前沿领域高质量数据集,涵盖自动驾驶、智慧医疗、智慧金融、机器人等行业应用中的核心AI需求。


Recommended>>
1、WF Quickspeed Future News GodGame vulnerability principle and hacking techniques analysis
2、Nissan Expands Collaboration with NASA on SelfDriving Technology Project
3、After Geelys acquisition of Proton Malaysia a delegation from Proton arrived in Hangzhou Bay for a visit
4、What do you know about EOS memory
5、How big is the space in the blockchain cloud services market

    已推荐到看一看 和朋友分享想法
    最多200字,当前共 发送

    已发送

    朋友将在看一看看到

    确定
    分享你的想法...
    取消

    分享想法到看一看

    确定
    最多200字,当前共

    发送中

    网络异常,请稍后重试

    微信扫一扫
    关注该公众号