These days, a business, not artificial intelligence, simply can not get money.
Large companies engage in artificial intelligence, big company executives quit venture to engage in artificial intelligence, artificial intelligence to make artificial investors investment, Intelligent Company undoubtedly become the current Internet Co in the pool of gold, no reliable company, as long as the reliable staff, can get a lot of money, enough to thump in artificial intelligence in the tide of a time.
But to be practical, the research directions of artificial intelligence in these technology companies are mainly three: deep learning, natural speech processing, machine vision. Deep learning is a required course for artificial intelligence. Natural Language Processing can directly apply the technology to the virtual intelligence assistant, and machine vision can be directly transformed into automatic driving technology.
Intelligent speakers equipped with virtual voice assistant has appeared for a long time, but from the perspective of consumer feedback, they do not feel the intelligent voice assistant is a smart thing, even now is intelligent speaker looks a little chicken ribs. And smart speakers although the market is hot, but in the consumer circle in different cold, automatic driving seems to be a investment ｏｒ ordinary consumers are optimistic about the project. Many people expect automatic driving to improve the traffic situation in the future, solve the current congestion and inefficient traffic problems, and the capital market is also particularly favored auto driving companies and drivers.
This morning, Momenta, a start-up automation company, announced a $46 million investment in the B round. The round of financing by Wei capital lead investment, Daimler group (Mercedes Benz parent company), Shun capital, Innovation workshop and nine venture capital with cast. In 2016, Momenta received A round of financing from Blue Lake capital holdings, innovation workshops and real estate funds. In early 2017, access to the capital for the lead A1 round of financing.
The company was founded in less than a year, into multiple rounds of funds, B round 46 million is indeed a small investment, especially the company has not yet any external output.
Come and dig down this company.
Momenta, a Beijing based auto pilot company, was established in 2016 September. Founder Cao Xudong previously worked at Microsoft Asia Research Institute of science and technology, and. The goal of the company is to Austria automatic driving of the brain, the core technology is driving the decision algorithm of environmental perception, deep learning of high precision map, based on. Products include different levels of self driving programs, as well as large data services derived from them.
That is to say, Momenta is an auto driving program company.
In contrast to other companies, Momenta's technological advantage lies in visual recognition. According to Cao Xudong introduction, Momenta has the world's top experts in depth learning, the most advanced framework in the field of image recognition Faster, R-CNN and ResNet author, ImageNet 2015, MS, COCO, Challenge more than 2015 titles. The team came from Tsinghua University, Massachusetts Institute of Technology, Microsoft Asia Research Institute and so on. It has deep technical accumulation and strong technical originality. The company's R & D. director is the inventor of Faster RCNN, the most widely used object detection framework in the world.
The advantage of such visual recognition technology enables the company to gain a greater advantage in the most basic sensory part of the autopilot program. Because from the current point of view, the visual recognition will become main part of the automatic driving scheme of laser radar sensing, compared to the beginning of automatic driving cars used in large quantities, the camera sensor is lower cost, the data is less than the laser radar surveying and mapping environment generate data is automatically driving ｏｒ driving auxiliary solutions in comparison the actual.
But the visual recognition of the difficulty lies in the visual recognition algorithm with high precision, only the optimization algorithm, the on-board computer to the camera images collected were identified and treated can be divided into driving area, pedestrian street, road traffic factor.
And algorithms are exactly what Momenta is good at.
But to realize automatic driving, it is impossible to have good perception ability. Decision and control are the most critical technical modules of automatic driving. In order to adjust the automatic driving decision algorithm, a large amount of actual road test data is needed to optimize the algorithm so as to realize the relatively intelligent and safe autopilot control.
But the accumulation of automatic driving data is a money burning work, small companies are difficult to achieve, so it requires huge financial and human support.
So, let's go back and see where the $46 million financing will go. Cao Xudong said: "the financing will be used in three aspects: 1. to strengthen the core competence of artificial intelligence, including big data, big calculation and outstanding talent AI; 2. product of the environmental perception of visual and high precision based on the technology of high frequency; 3. R & D just need the scene L4 unmanned technology."
At present the houses are in the accumulation of automatic driving technology, talent is the most expensive investment, because each have roped in automatic driving technology talents, leading talent in this area of revenue, many large companies are even willing to throw billions of dollars, as is the acquisition of several technical personnel set up automatic driving small team.
On the other hand, for the automatic driving of entrepreneurial companies, capital will not allow long-term investment in unproductive consumption, everyone thinks the future of automatic driving can make money, but the automatic driving what can earn much money, the money in the autopilot is not yet universal how to earn, now is still a problem, so start-up companies still need to own the technology has produced