"You mean Baidu's result is related to Meng we are going to meet in the next two days?" Jeff's heart tightened. His only mission when he came to Australia this time was to recruit Meng Fanqi. It couldn't be such a coincidence, right? "How long has it been since he created DreamNet and Generative Adversarial Networks? One or two months, how long is it enough to achieve such good results in a different field?"
This is the thinking inertia of ordinary people. If a person has just accomplished one or two big things, then usually no one will think that he is also preparing other things of the same level.
"Furthermore, there are no white people as co-authors in his generative adversarial network article, and my information description does not mention any white institutions." Alex was shocked by DreamNet's amazing surpassing of himself, so he read it carefully Pass the paper on generative adversarial networks published by Meng Fanqi. “It shouldn’t have anything to do with whiteness.”
Alex read this paper carefully, word for word, and understood it very carefully.
"But how to explain the excellent performance of Baidu's new detection algorithm?" Hinton stood aside and thought for a while, "If it is based on deep learning, the current detection algorithm is nothing more than two points, the backbone neural network and the detection head part filter categories.”
"This speed improvement must mainly come from innovation on the detection side, otherwise it will not be enough. But this level of performance leap is absolutely inseparable from the neural network as the backbone." Hinton has a very sophisticated eye. , basically guessing the source of improvement of the YOLO algorithm.
"As far as the performance of DreamNet is concerned, whiteness is absolutely inseparable from Montenegro." Hinton's perspective is more from the analysis of the technology itself. "The timing is also quite coincidental. It may be that Baidu's detection innovation and Meng Fanqi's backbone breakthrough are powerfully combined."
Hinton's thoughts were still conservative. He never expected that two amazing technological breakthroughs were made by the same person.
"However, Meng also had quite good results on the test track of this competition." Ilya added.
The most popular IMAGENET series competition is the classification of more than 10 million images. The number of participants in the detection track is actually much smaller, and it has received very little attention.
It is a pity that this year's testing event does not compete with speed, so a few people cannot use the results of this track to make further speculations on the matter.
"He won't be poached by Baidu, right?" Jeff frowned slightly, feeling bad in his heart. If it is really signed away first, it will be troublesome.
When I poached the three Hintons last year, I was just a little bit behind, almost being beaten by Baidu. Baidu was quite fast in this regard.
The last time multiple technology companies bid, Baidu was not only the first one to contact Hinton ahead of him, but he was also the one who fought with him to the end.
If it weren't for the Hinton trio's preference for Google, it's still unclear whether they would be able to defeat whiteness in the end.
Could it be said that the way of heaven is reincarnated, and now it is my turn to lose to Baidu who has a geographical advantage?
"Alix, we have an appointment this afternoon, right?" Jeff asked, "Is he in Australia now?"
"Here we go, this morning's flight. I asked him two days ago if he wanted us to arrange a pick-up, and he said it's no trouble. I'll see you in the afternoon."
"That's good..." When he heard that Meng Fanqi had arrived in the morning, Jeff felt relieved again. As long as you are willing to talk to yourself, it means that Baidu has not succeeded yet.
I'm just afraid that this kid, a poor undergraduate student who has never seen much of the world, will be abducted by Baidu and spend millions to sign the contract without giving him a chance to make a bid. That would be too frustrating.
As long as he allows himself to speak, Jeff is absolutely confident that he will win over the other party at the negotiating table.
With a top-notch salary budget, a relaxed working environment, and the AI monument Hinton, Jeff really couldn't imagine the reason why a college student who was engaged in AI as an undergraduate would reject him.
Meng Fanqi had deliberately not left his contact information, which originally made Jeff think it was a good thing.
An unknown person who has never been published. If they don't want to be disturbed, other companies should not know about his situation except Alex, who has exchanged emails.
Therefore, Jeff did not rush too much in arranging the meeting time. He could wait for more than 20 days.
"It must be that damn generative adversarial network paper." Jeff said with some annoyance. This is a problem of geographical location.
If Meng Fanqi was in the United States, he could fly over on the same day and win it directly without delaying a day.
But when he wanted to go to Yanjing, he still hesitated. He felt that the other party didn't want to contact any company, so he was taking a chance.
"Just wait until we meet in the afternoon to talk. Jeff, you don't have to think so much, it's useless." Hinton knew that Jeff was more outgoing in private and did not hide his emotions.
The negotiation at two o'clock in the afternoon was held at the venue of the International Computer Vision Conference.
Unlike what everyone imagines, large conference rooms or private clubs, many foreign AI-related negotiations and talks are actually conducted in the process of international conferences.
A few big names in the industry or a certain field entered the small meeting room in twos and threes, or the room where afternoon tea was provided, or even the smoking room.
In such a seemingly hasty and casual scene, the future pattern of many technologies and enterprises was finalized.
This method is actually more to Meng Fanqi's liking.
At noon on December 5, 2013, when the morning meeting ended, everyone was chatting and bragging while eating.
Some sharp-eyed people have already spotted the fleeting figures of Jeff and Hinton.
"Did I read that correctly? Is that Hinton?" A young researcher said to his companion in surprise, "I heard that Hinton rarely participates in conferences held outside the United States and Canada due to physical reasons. Well, even in the United States and Canada, we still have to look at the location and climate, why did we end up in Sydney?"
"The one next to him is Jeff. This big guy rarely attends conferences." After all, Jeff is not a researcher or scholar in the field of AI, but as a core technical staff of Google, he participates in conferences and publishes papers. Generally speaking, it is not his turn to take action personally.
"What kind of wind blew these two great gods to Xini?" The two people's low-key whereabouts still caused a small commotion and discussion in the venue.
Most people kept a polite distance and did not continue to pay too much attention or discuss.
But consciously or unconsciously, they were paying attention to that direction, that small conference room. I hope to know who it is that wants to have direct dialogue with these two giants, making it mysterious.
It’s impossible for the four people from Google Brain to come all the way from the United States to the venue in Sydney to hold an internal meeting, right?
