Static Gestures Dataset

The Whole Body Static Dataset showcases a diverse set of still body poses, including motions in both the upper and lower body. As opposed to our Animated Dataset, the diversity of poses and environments are increased more than 100x in this static dataset. This dataset is well suited for ML tasks such as pose estimation, whole body segmentation, and basic gesture recognition.

Synthesis AI provides programmatic generation of vast amounts of photorealistic image data with an accompanying set of never-before-available rich 3D labels. Learn more about our platform.

*Free, Open Non Commercial License

ABOUT THE SET

Specifications

10k images
100 diverse identities spanning gender, skin tone, and age. 1024 x 1024 RGB images
Rich set of pixel-perfect labels including segmentation maps, depth, surface normals, and 3D landmarks
Variation of expressions from neutral to medium
Variation in eye gaze spanning 35° vertically and 45° horizontally
Diverse hairstyles, colors, and facial hair
Diverse static poses
Accessories: hats, glasses, headphones, and face masks
Variation in outfits
Variation in background over a wide range of lighting and confounding elements
Variation in camera angles spanning 30° in all directions around the person

KEY MACHINE LEARNING TASKS

This data can be used to train...

Pose Estimation

Wide variation in camera perspectives challenge models in being robust to camera orientation.
Granular subsegmentation maps include X regions such as “ear_left”, “eyelashes”, “eyelid”, “hair”

Gesture Recognition

Over 30 gestures with hand motion, 50 gestures featuring whole body motions.
Wide variation in skin tone and lighting.

Whole Body Segmentation

Large segmentation areas distributed over the body, including the arms, hands, background, hair, glasses, skin, and more.
Granular sub segmentation maps include 90 regions such as “neck”, “torso_upper_right”, “hand_right”, “finger1_mid_left”.

To download the dataset, fill out the form.