FLIC - Frames Labeled In Cinema¶
Collection of 5003 images automatically from popular Hollywood movies. It contains annotations of upper body joints only. It has 3987 images for training and 1016 for testing.
Use cases¶
Human body joint detection.
Properties¶
name
: flickeywords
: image_processing, detection, human_pose, keypointsdataset size
: 300,3 MBis downloadable
: yestasks
:- keypoints: (default)
primary use
: image classificationdescription
: Contains keypoint coordinates (upper body joints only) and bounding boxes of the torso for body joint detection.sets
: train, testmetadata file size
: 582,0 kBhas annotations
: yeswhich
:- image filenames
- upper body joint coordinates
- torso bounding box
Metadata structure (HDF5)¶
Task: classification¶
/
├── train/
│ ├── image_filenames # dtype=np.uint8, shape=(3987,108) (note: string in ASCII format)
│ ├── movienames # dtype=np.uint8, shape=(3987,31) (note: string in ASCII format)
│ ├── width # dtype=np.int32, shape=(3987,)
│ ├── height # dtype=np.int32, shape=(3987,)
│ ├── torso_boxes # dtype=np.float, shape=(3987,4)
│ ├── keypoints # dtype=np.float, shape=(3987,11,3)
│ ├── keypoint_names # dtype=np.uint8, shape=(11,15) (note: string in ASCII format)
│ ├── object_fields # dtype=np.uint8, shape=(5,16) (note: string in ASCII format)
│ └── object_ids # dtype=np.int32, shape=(3987,5)
│
└── test/
├── image_filenames # dtype=np.uint8, shape=(1016,108) (note: string in ASCII format)
├── movienames # dtype=np.uint8, shape=(1016,31) (note: string in ASCII format)
├── width # dtype=np.int32, shape=(1016,)
├── height # dtype=np.int32, shape=(1016,)
├── torso_boxes # dtype=np.float, shape=(1016,4)
├── keypoints # dtype=np.float, shape=(1016,11,3)
├── keypoint_names # dtype=np.uint8, shape=(11,15) (note: string in ASCII format)
├── object_fields # dtype=np.uint8, shape=(5,16) (note: string in ASCII format)
└── object_ids # dtype=np.int32, shape=(1016,5)
Fields¶
image_filenames
: image file path + nameavailable in
: train, testdtype
: np.uint8is padded
: Truefill value
: 0note
: strings stored in ASCII format
movienames
: name of the movie where the image was taken fromavailable in
: train, testdtype
: np.uint8is padded
: Truefill value
: 0note
: strings stored in ASCII format
width
: image widthavailable in
: train, testdtype
: np.int32is padded
: Falsefill value
: -1
height
: image heightavailable in
: train, testdtype
: np.int32is padded
: Falsefill value
: -1
torso_boxes
: torso bounding boxavailable in
: train, testdtype
: np.floatis padded
: Falsefill value
: -1note
: bbox format [x1,y1,x2,y2]
keypoints
: body joint coordinatesavailable in
: train, testdtype
: np.floatis padded
: Falsefill value
: -1note
: keypoint format [x1,y1,is_visible]
keypoint_names
: body joint nameavailable in
: train, testdtype
: np.uint8is padded
: Truefill value
: 0note
: strings stored in ASCII format
object_fields
: list of field names of the object id listavailable in
: train, testdtype
: np.uint8is padded
: Truefill value
: 0note
: strings stored in ASCII formatnote
: key field (field name aggregator)
object_ids
: list of field idsavailable in
: train, testdtype
: np.int32is padded
: Falsefill value
: -1note
: key field (field id aggregator)