Data labeling changes from offline to online.
It is learned that the AI data service platform My Neighbor Totoro Data has recently received a 33 million yuan Pre-B round of financing led by KIP China and Jinshajiang Venture Capital. This round of financing will be mainly used for market expansion, automated labeling research and development, AI research and development investment. The company has previously received multiple rounds of investment from Jinshajiang Venture Capital, Jiuhe Venture Capital, Wudou Venture Capital, Zhenshun Fund, and Cloud Angel Fund.
Data, computing power, and algorithms are the three elements of the development of artificial intelligence. The development of artificial intelligence technology has created a market demand for data labeling. In order to obtain massive and diverse data, the early data labeling teams mostly carried out offline, employing human resources in outsourcing, and borrowing many human resources for data collection and labeling, which made the industry’s labor-intensive characteristics particularly prominent. The double increase in data demand and data quality requirements, high labor costs, low service efficiency and low quality have become the main problems in the current industry development.
My Neighbor Totoro Data has shifted the data labeling work from offline to online based on the difficulties faced by the above industries. By building an AI data platform, it attracts online users to participate in data collection and labeling to reduce labor and site costs. Through continuous optimization of data annotation tools, automatic annotation, and pre-annotation of data, the efficiency and quality of data annotation is improved.
My Neighbor Totoro data platform was launched in 2016, and its service areas cover computer vision, speech semantics, autonomous driving, and traditional industry AI + .
From the functions of the data platform , the company has implemented AI automatic labeling. Based on the accumulated labeling items and data, it develops tools for face recognition, image semantic segmentation, OCR, Scenario data such as speech recognition and automatic driving are pre-labeled, and then the tasks are subcontracted to the labeling personnel. Therefore, targeted training and iteration of AI can be achieved. In addition, the platform’s acceptance function allows customers to follow up the data production process remotely, improving the customer experience.
From the perspective of tagger management and tagging quality control , the platform acquires taggers in a way that attracts traffic, and can obtain more human resources than traditional offline tagging scenarios. In addition, taggers receive training on the platform, assign different tasks based on the completion effect and proficiency of the tagging, and get paid according to the actual results. These control measures can effectively improve data quality, and at the same time, make service offers 30% cheaper than traditional offline models.
Currently, the platform has about 5 million registered users and more than 2,000 labeled teams. It has served hundreds of Internet and technology R & D head companies, including Baidu, Ali, Samsung, JD.com, Toutiao, Xiaomi, Ping An, Companies such as China Mobile have been using the services provided by My Neighbor Totoro for a long time.
The epidemic has also promoted the change of data labeling from offline to online. Currently, many offline service AIData companies are facing downtime, while Chinchilla Data’s platform data delivery has not been affected, and a large number of AI-specific application data developed in response to the epidemic have also been launched.
In terms of the overall market, the data indicates that the industry started late in the country. The industry’s representative companies include Appen with a market value of more than $ 2.8 billion, Amazon’s AMT, Scale AI with a value of $ 1 billion, and recently completed $ 25 million Labelbox and other round financing.
Kong China Investment Director Lin Yong said: In the process of landing artificial intelligence in various vertical industries, the demand for data services will continue to erupt. As a domestic leading artificial intelligence data service provider, Totoro is recognized by many high-quality customers at home and abroad, while the supply side has flexible and flexible organizational capabilities, and hopes to provide better infrastructure services for companies with AI business needs.
Jinsha Jiang Venture Partners partner Chen Jiejie said: My Neighbor Totoro adheres to technology first, and continues to deepen its first-mover advantage in technology iteration through technology iterations. It ranks domestically in the accuracy, speed, and intelligence of data services. Jinshajiang Venture Capital continues to be optimistic about the core capabilities of the My Neighbor Totoro team to provide artificial intelligence data services, is optimistic about the development prospects of artificial intelligence in China, and continues to pay attention to and actively arrange investment opportunities for artificial intelligence landing in various vertical industries.