Table 3.

Challenges, limitations, and future directions of different computer vision sensing and data processing approaches.

	Current approaches / Problems	Future directions / Improvements/ Challenges
Sensors	Current approaches	Future directions
	Mostly 2D single cameras	Multiple 2D cameras
	Problems	3D (depth) sensors
	Only 2D information	Pressure mat sensors
	Occlusions	Improvements
		3D information
		Less occlusions
		More information due to multi-sensory integration
Data	Current approaches	Future directions
	Small datasets	Collect more data
	Problems	Make it publicly available
	Not enough data to employ deep learning methods	Make use of home videos
	Not publicly available - no benchmarking possible	Employ DL methods
	Incorrect or incomplete data in some cases, e.g., inaccurate outcome labelling due to lack of longitudinal studies, the inclusion of incorrect age-specificity cases, use of low-inter-rater agreement or small rater-group or lack of experienced raters in data labelling, disorders or gender misrepresentation	Make use of transfer learning (e.g., Tan et al., 2018)
		Challenges
		Need to solve anonymisation issue (automated techniques for face detection and replacement can be applied)
		Development of methods which can cope with different light conditions, resolution, frame rate
Body areas of interest	Current approaches	Future directions
	Mostly movement of arms, legs, head	Hand, fingers, feet
	Problems	Eye movement data
	Incomplete information of full-body movement	Mimic
		Challenges
		Integration and analysis of multimodal information
Motion tracking	Current approaches	Future directions
	Mostly in 2D space	Full-body tracking in 3D using well-established methods in DL (e.g., DeepLabCut and OpenPose frameworks)
	Problems	Challenges
	Only 2D information	DL methods need to be adapted to infants
Motion encoding	Current approaches	Future directions
	Conventional features based on: displacement, distance, velocity, acceleration, speed, and time	Motion encoding using well-established methods from robotics:
	Problems	Dynamic Movement Primitives (e.g., Ijspeert, Nakanishi, Hoffmann, Pastor, & Schaal, 2013), Gaussian Mixture Models (e.g., Calinon, 2016; Khansari Zadeh & Billard, 2011), Probabilistic Movement Primitives (Paraschos, Daniel, Peters, & Neumann, 2018)
	Only 2D features	Learn features from expert knowledge during observation (e.g., Silva et al., 2018, 2019)
		Improvements
		3D features
		New motion encoding and features
Classification algorithms	Current approaches	Future directions
	Conventional ML methods, e.g., SVM, Decision Trees, Neural	Employ ANN, DL
	Networks, Hidden Markov Models	Employ Interactive Machine Learning (learning with feedback)
	Supervised learning without feedback during learning	Improvements
		Better models with more accurate predictions
		Challenges
		More data is needed