Video-based activity recognition

An apparatus for processing one or more frames, comprising:

• a storage device;
• and
• a processor coupled to the storage device, the processor configured to:
• extract, using a first machine learning model, first one or more features from a first frame, second one or more features from a second frame, and third one or more features from a third frame, the first one or more features, the second one or more features, and the third one or more features being associated with a person driving a vehicle;
• store the first one or more features, the second one or more features, and the third one or more features in the storage device;
• obtain, from the storage device, the first one or more features, the second one or more features, and the third one or more features;
• remove features from the storage device based on storing the third one or more features in the storage device;
• process, using a second machine learning model, the first one or more features, the second one or more features, and the third one or more features;
• and
• determine, based on processing of the first one or more features, the second one or more features, and the third one or more features using the second machine learning model, at least one activity associated with the person driving the vehicle.