***Please visit the database here.

***If you’d like to contribute or collaborate on this database, please submit your contact information here.

Developed by Su Lei, Kelsie Lam and Jonathan Gratch.

If you publish work based on this database, please cite, “Lei, Lam, Gratch, AAAC 2019 Commercial Product Survey, Association for the Advancement of Affective Computing, https://aaac.world/productdb/, 2021”.

A survey was distributed between March and May 2020 in the AAAC community. A total of 380 unique visitors attempted to answer our survey, which resulted in 125 valid responses. A response is considered valid only if the participant reported at least one affect computing product.


Participants are from 20 countries in 3 languages, 45% in English, 25% in Japanese and 30% in Chinese. As shown in the Figure above, most representative countries are China (31%), Japan (27%), US (14%), UK (7%), and France (3%). In terms of sectors, 45% of the participants are associated with educational institutions, 29% of them are full-time students, and 22% of them are employed by private companies. Specifically, 78% of participants are practicing fundamental research––42% of them are advancing capabilities of algorithms and 36% of them are advancing understanding of people. Besides fundamental researches, 34% of them are working in the field of behavioral analytics and 20% of them are working on projects related to mental health care. In terms of employment role, 86% of the participants identified themselves as researchers, 30% of them considered themselves as educators, 14% as practitioner (e.g., software engineer) and 14% as managers. Note participants are allowed to choose multiple sectors, industry and roles.

Participants reported 151 entries of emotion recognition products in total. 18% of participants indicated association with the products they reported. 29% of them tested or used the products that they reported. 34% of the entries are products participants only heard about. Among all the reported entries, we kept only entries that could lead us to an identifiable product or service (e.g entries are removed if participants simply indicated they have heard about products of facial expression recognition without specifying which one), and we removed repeated mentions and projects that are only within the scope of academic lab research.

The survey itself leads us to 50 products in total. During the post analysis of the survey, we have expanded the database to 111 products at the time of writing this post.

We would like to thank everyone who participated in the survey, and special thanks to Dongmei Jiang and Kazunori Terada for translating the survey to Mandarin and Japanese.