The crowds arrive daily at Flushing Meadows Park in Queens. It’s the US Open and all eyes are on the top tennis players in the world. With 254 matches just in the men’s and women’s singles tournaments, it’s a lot to keep up with, but behind the scenes, IBM technology is doing just that.
So this gets updated at the end of that points. We see how Watson evaluate at that point. With so much data to sort through, IBM is employing its artificial intelligence software Watson to help. In addition to delivering scores, Watson also scans game footage. It listens to the roar of the crowd.
We’ve trained it to look for fist pumps and gestures of the players as well as to look for facial expressions. All of that contextual data is used to measure the momentum of any given match in update fans after each shot. A few years ago, we didn’t really perceive video as data, but it really is. There’s 17 courts, so it’s a lot of action, it’s a lot of video for the editors to digest, so now that you can quickly identify the most exciting moments and turn it around in near real time for fans. These highlights aren’t just for fans. Coaches also use them to prepare players.
In the past, it was a really labor-intensive process where the matches had to be manually tagged, and by tagging, I mean somebody goes through the video footage and makes a note of every forehand, every backhand, every unforced error. It takes hours and hours. Watson’s artificial intelligence processes and indexes video in minutes, freeing up valuable time for personalized coaching. If the coach has a certain pattern in mind that they think will be effective in the next match, we can generate a playlist that shows the player executing those patterns and really reinforce and ingrain that in their mind.
Players’ facial expressions and body language can also be correlated to their performance. Even the way their grunting, there’s certain trends and you know sometimes, they are counterintuitive, some players play better when they get a little bit angry, when they get a little upset. With the help of artificial intelligence, perhaps getting a little upset will lead to more upsets on the court.
Tina Turin, VOA News, New York
1. Players’ facial expressions and body language can also be correlated to their performance.
facial expression 面部表情
To help answer specific questions about presenting a speech, we will consider three major categories of nonverbal behavior that affect delivery: body language, eye contact, facial expression.
2. With the help of artificial intelligence, perhaps getting a little upset will lead to more upsets on the court.
with the help of 在……的帮助下
Only I struggle, when I struggle, with the help of Him that knows no struggle.
分数有变化时，这里要更新，所以我们能看到沃森（人工智能程序）的评估方式 。有太多数据要理清头绪，所以IBM开始启用人工智能软件沃森来助一臂之力 。除了可以显示分数外，沃森还能显示分数之外，沃森还能用于回顾游戏片段，并记录观众的呐喊声 。
我们对沃森进行了训练，让它可以识别选手的拳头和姿势，还能识别面部表情 。每次击球后，这些数据都可以用来衡量比赛当天的动能 。数年前，我们不会把视频作为数据看待，但现在看，视频确实是数据 。有17个场地，有很多的动作要捕捉，编辑有很多视频要消化，而有了人工智能后，就能锁定最激动人心的时刻，然后调取出接近实时的时刻给球迷们看 。这样的亮点时刻不仅是为球迷准备的，教练也会借助这个来让球员做好上场前的准备 。
以前的工作量很大，所有比赛都用手动调 。手动调是指要把整个录像片段看一遍，要记录每一次正手、反手、非受迫性失误 。要花很多个小时才能做完 。沃森这个人工智能软件可以按分钟来处理视频并贴上标签，这样就可以为个性化指导腾出宝贵的时间 。如果教练心中有了某个模式，认为这个模式有助于在下一场比赛中取胜的话，就可以生成一个列表，从中看到选手在操练这种模式，并在脑海中加强巩固 。
选手的面部表情和肢体语言与他们的场上表现息息相关 。他们自言自语的方式，也是有倾向的 。有时候会与直觉相悖，因为一些选手在略微生气或者躁动的时候反而会发挥的更好 。在人工智能的帮助下，或许些许的躁动会制造场上更大的兴奋 。
这里是蒂娜 都灵(Tina Turin)从纽约发回的报道 。