Case Study: UK-Based Wearables Development Company achieves ML-based emotion recognition for fan wearables with Intetics

Machine Learning Algorithm to Recognize Human Emotions

UK-Based Wearables Development Company needed to add an emotional-response feature to its sports-wearable product so fans could share reactions during live events. The company engaged Intetics to develop a supervised machine learning algorithm for emotion recognition using biosensor data collected at games, working to meet a strict delivery timeframe.

Intetics implemented a workflow including a visual labeling tool, preprocessing to filter and align sensor signals, segmentation (1–3s windows), feature extraction, PCA to reduce dimensionality by about four times, and a 70/30 train/test split to train and validate the classifier. The resulting Intetics-built algorithm proved that emotion recognition from biosensor data is feasible, enabled the new product feature to be added (including real‑time on‑device classification), and helped boost customer loyalty to the product.