Decoding consumer’s face value

Vijay C Roy

Tribune News Service

Chandigarh, September 19

Companies will soon be able to read the mind of consumers and accordingly plan their marketing strategies through the intelligent face-tracking software developed by a team of IIT Ropar.

Even before the technology came out of the laboratory, a global electronic giant has bought it. The firm would modify and use it to map the mood of people captured in selfies, said Dr Abhinav Dhall, an assistant professor in IIT Ropar. Dhall did not disclose the details of the technology transfer, citing a non-disclosure agreement with the company.

There could be various applications of this artificial intelligence-based software as it will analyse hundreds of faces in a group photo and provide “mood score”, which is data related to potential group behaviour, he said. Companies can also obtain real-time data pertaining to the consumer profile such as age and gender. Dhall and his team in the Department of Computer Science and Engineering, IIT Ropar, have developed the technology.

The software captures three interesting types of information in an image; facial expression, upper body gestures and information related to the surrounding, he said. The research has been made in collaboration with researchers from the University of Oulu, Finland and University of Trento, Italy.

The software’s aim is to assess the overall mood of a group from a single shot, Dhall said. By looking at a sequence of frames in a video, it could even gauge the mood of a crowd in realtime. “If the mood score goes down over the time, we can assume that the group is getting angry,” he said.

Giving details of the usage of the technology for selfies, he said, “In the era of social media, we post hundreds of images in different moods under different scenarios. It will be interesting to analyse the overall moods.” This technology is already sold and the same will be used by selfie-lovers to select their pictures in right mood.

The IIT-Ropar has also signed an agreement with a company to study employee behaviour. Though the software could be used officially to improve HR practices of the company, experts say that it might be invasion to the privacy of the staff.

The software can be quite useful by the law-enforcement agencies in pre-empting the mob behaviour and take corrective measures in advance to avoid precipitation of any violence. Dhall’s team is developing an app to detect person under intoxication and substance abuse through their facial expressions. The software cannot only detect but also disclose vital data such as duration of substance abuse and the possibility of person’s recovery.

Its applications

• Launching products based on consumer behaviour
• Formulating HR strategies in line with employees’ efficiency
• Keeping best-mood selfies and deleting others
• Assessing mob behaviour