Facial recognition is quickly becoming the most common form of biometrics. It’s also the most likely form of biometrics you’ve already experienced (you probably use it multiple times a day to unlock your phone). Facial recognition is useful, but it’s important to understand that it is quite different to the facial expression analysis that we use in our research lab.
What is the difference? Both involve scanning the face and identifying particular features and dimensions. For instance, the width of your mouth or the length of your nose. Facial recognition uses information like this and matches it against information stored in its database for the purpose of verifying their identity. Facial expression analysis, meanwhile, is more focused on what expressions the person is displaying (e.g., smiling, frowning), using information on how the facial features are changing over time to determine the likelihood of an expression being displayed. For instance, if the width of the mouth grows, the probability of the person smiling goes up. This makes facial expression analysis much more relevant for market research, because it allows us to automatically determine how a participant is responding emotionally to a product or service. Facial expression analysis is not concerned about the identity of the person, it simply cares about what expressions they are displaying. It’s less about the who, and more about the what and why.
Because the identity of the person is not the focus, facial expression analysis does not come under the same sort of scrutiny as facial recognition (which has come under fire for its potential misuse in surveillance). Nonetheless, both these forms of biometrics require responsible data collection, which is something we take seriously as researchers in an emerging field of technology. All participants have a right to know what data is being collected and where it will end up.
Our software provider, iMotions, recently published a blog post going into more depth on this topic, highlighting some of the ways that we as researchers ensure data is handled responsibly and transparently:
A key tenet of research ethics is that if a study is not conducted ethically, it is not considered scientifically valid, just as if there were any other methodological flaws with the research. Facial expression analysis has enabled us to conduct studies that would have never been possible, and we ensure that all our results are delivered both insightfully and responsibly.