Spectee launches a new beta platform ‘SIGNAL’, providing solutions to meet various AI needs ~ Commencement of real-time video analysis for weather camera
Spectee Inc. (Chiyoda-ku, Tokyo) provides data analysis services using artificial intelligence (AI) to companies that require AI development in areas such as image analysis, natural language analysis, voice and speech recognition. SIGNAL enables companies to use and develop AI quickly, easily, and cost-effectively.
Although many companies intend to implement AI, the high cost, lack of data for machine learning, and long development process pose as a barrier and challenge. They could not immediately implement AI in their business and end up having do without it.
AI development platform SIGNAL is a service which addresses these issues and makes it possible for companies in need of AI to implement it quickly at a lower cost.
For example, companies can utilize SIGNAL to identify certain objects within an image. By analyzing about 10 different raw images necessary for machine learning, SIGNAL is able to reach the point that images can be processed with high accuracy.
AI Development Platform (SIGNAL)
Since the launch of Spectee SNS alert service in 2014, over 220 clients including news media organizations, government administration office, municipality, private enterprises, etc., are already subscribed to our service. Other services such as AI announcer “Araki Yui”, AI integrated tourist information service “Localive”, AI automated speech/voice recognition and transcription service “VoiceClip” and so on. With the focus centered on AI image analysis, Spectee has conducted intensive research and developed various types of AI.
In addition, in 2019 March, technology company, NVIDIA, certified Spectee as a partner under the “NVIDIA Inception Program”. NVIDIA highly regarded Spectee AI development was well acknowledged by NVIDIA in various aspect.
AI development platform SIGNAL is a service platform that introduces the use of AI which was brought together by the technologies and know-how that Spectee has cultivated so far. It supports more companies at a lower cost, faster and easier. Through SIGNAL, Spectee will continue to support various companies in AI development.
Spectee, Fuji Television and the Japan Weather Association jointly launched a demonstration test of clothing information using the AI analysis of a live camera. (Broadcasting on the weather section of “Mezamashi Television” from April 1.
In recent years, an increasing number of days with abnormal weather condition have been observed. Colder and warmer days have occurred outside of each season. In a survey conducted by the Japan Weather Association (JWA), the following question was investigated.
Question: Please select the purpose and reason for checking the weather forecast in the past one week.
The results showed that the most popular reason was that people used the weather forecast to determine if there was a need to bring along an umbrella (69.3%). The next most popular reason was to determine what clothes to wear when they are heading out (46.5%).
From these results, the JWA found out that statistics related to clothing and personal belongings on the actual day is important to those who uses the weather forecast. Spectee realizes there was a demand for real-time weather information analysis by utilizing the footage taken via weather cameras already available on the streets.
In cooperation with the JWA and Fuji Television, Spectee implemented an AI video analysis tool based on the AI development platform 'SIGNAL' in about 2 months. Since April 1, 2019, Spectee provided detailed and valuable information on the weather such as the percentage of people wearing coats, long vs short sleeve shirt ratio and umbrella usage rate. "AI Weather" was implemented and broadcasted live.
Attire Distinguishing Image
Image of Fuji TV series "Mezamashi (Alarm Clock) TV"
[Experimental demonstration of AI analysis of clothing information using camera image]
Fuji TV series "Mezamashi (Alarm Clock) TV": At around 6:50 am in the weather corner (Nationwide)
Broadcasting period: April 1-30, 2019 (Scheduled) *The end date may be changed.
*About the survey (Conducted by JWA as part of its service)
Survey period: March 9~12, 2019
Target Audience: About 800 people living in Tokyo, Kanagawa, Saitama and Chiba
Survey Method: Internet survey
Others: Age and gender are based on demographics, multiple answers were accepted