Self-touch and other spontaneous behavior patterns in early infancy

Khoury, Jason F.; Popescu, Sergiu T.; Gama, Filipe; Marcel, Valentin; Hoffmann, Matej

doi:10.1109/icdl53763.2022.9962203

Cited by 1 publication

(2 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although automated pose estimation seems to be generally good (see Fig. 1 a-c), in a previous study 34 , we encountered some limitations of current pose estimation methods in infants. For example, OpenPose 17 was observed to struggle with some camera angles, body postures, and keypoints.…”

Section: Introductionmentioning

confidence: 81%

“…With this reference distance, we normalize the errors between recordings to compensate for different camera settings and infants of different ages and with different body sizes. Others and ourselves have previously used a similar strategy in a similar context 18,34 . Unlike OKS, such errors are easier to grasp as they directly relate to a measurable part of the body of the infant.…”

Section: Neck-midhip Errormentioning

confidence: 99%

See 1 more Smart Citation

Automatic infant 2D pose estimation from videos: comparing seven deep neural network methods

Gama,

Mísař,

Navara

et al. 2024

Preprint

View full text Add to dashboard Cite

Automatic markerless estimation of infant posture and motion from ordinary videos carries great potential for movement studies “in the wild”, facilitating understanding of motor development and massively increasing the chances of early diagnosis of disorders. There is rapid development of human pose estimation methods in computer vision thanks to advances in deep learning and machine learning. However, these methods are trained on datasets featuring adults in different contexts. This work tests and compares seven popular methods (AlphaPose, DeepLabCut/DeeperCut, Detectron2, HRNet, MediaPipe/BlazePose, OpenPose, and ViTPose) on videos of infants in supine position. Surprisingly, all methods except DeepLabCut and MediaPipe have competitive performance without additional finetuning, with ViTPose performing best. Next to standard performance metrics (object keypoint similarity, average precision and recall), we introduce errors expressed in the neck-mid-hip ratio and additionally study missed and redundant detections and the reliability of the internal confidence ratings of the different methods, which are relevant for downstream tasks. Among the networks with competitive performance, only AlphaPose could run close to real time (27 fps) on our machine. We provide documented Docker containers or instructions for all the methods we used, our analysis scripts, and processed data at https://hub.docker.com/u/humanoidsctu and https://osf.io/x465b/.

show abstract

Section: Introductionmentioning

confidence: 81%

Section: Neck-midhip Errormentioning

confidence: 99%