TABLE III. Overview of the Benchmark Challenges Currently Implemented in the Gaitmap-Challenges Package.

Processing Step	Dataset	Description
Stride Segmentation	EgaitSegmentationValidation2014 [42]	Algorithms need to identify individual strides based on the definition by Barth et al. [42] during supervised 4 × 10 m gait tests and unstructured simulated free-living activities. The sensors are mounted laterally at the shoes.
	EgaitSegmentationValidation2014 [42] (relabeled)	Same as the previous challenge, but in addition to straight strides the algorithms also need to identify strides during stair walking and turns.
	SensorPositionComparison2019 [55]	Algorithms need to identify individual strides based on the definition by Barth et al. [42] during various walking tasks in the laboratory. Two synchronized IMU sensors are mounted on the instep of the shoe.
Spatial Parameter	EgaitAdidas2014 [53]	Algorithms need to estimate the stride length of strides performed at different speeds within the capture volume of a MoCap system. Two sensors are mounted laterally on the shoe and predefined stride borders and events are provided.
	EgaitParameterValidation2013 [39]	Algorithms need to estimate the stride length of strides performed by geriatric patients while walking over a GAITRite carpet. Two sensors are mounted laterally on the shoe and predefined stride borders and events are provided.
	SensorPositionComparison2019 [55]	Algorithms need to estimate the stride length of strides performed at different speeds within the capture volume of a MoCap system. Two synchronized IMU sensors are mounted on the instep of the shoe and predefined stride borders and events are provided.
Full Pipeline	SensorPositionComparison2019 [55]	Algorithms need to estimate the mean and standard deviation of the stride length, gait velocity, and stride time of each of the performed gait tests. Only the raw IMU data of the two sensors mounted on the insteps of the shoes is provided.
	KlugeParameterValidation2017 [56]	Algorithms need to estimate the mean and standard deviation of the stride length, gait velocity, and stride time of each of the performed 4 × 10 m tests. Only the raw IMU data of the two sensors mounted laterally on the shoes is provided.

The Two Challenges With Descriptions Highlighted in Bold are Presented in More Detail in This Manuscript. The Dataset Column Provides the Class Name of the Dataset Within the Gaitmap-Datasets Package.