UCF Sports Action

UCF Sports dataset consists of a set of actions collected from various sports which are typically featured on broadcast television channels such as the BBC and ESPN. The video sequences were obtained from a wide range of stock footage websites including BBC Motion gallery and GettyImages.

The dataset includes a total of 150 sequences with the resolution of 720 x 480. The collection represents a natural pool of actions featured in a wide range of scenes and viewpoints.

Use cases

Human action recognition in videos.

Properties

  • name: ucf_sports
  • keywords: image_processing, recognition, detection, activity, human, single_person
  • dataset size: 1,8 GB
  • is downloadable: yes
  • tasks:
    • recognition: (default)
      • primary use: action recognition in videos
      • description: Contains videos and action label annotations for action recognition
      • sets: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
      • metadata file size in disk: 1,0 MB
      • has annotations: yes
        • which:
          • activity labels for each video.

Metadata structure (HDF5)

Task: recognition

/
├── train01/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(6548,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(6548,74)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(103,24)    (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(6548,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(103,127)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(103,127)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(103,127)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,15)
│
├── test01/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(3032,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(3032,76)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(47,24)     (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(3032,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(47,144)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(47,144)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(47,144)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,7)
│
├── train02/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(6529,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(6529,76)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(103,24)    (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(6529,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(103,144)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(103,144)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(103,144)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,15)
│
├── test02/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(3051,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(3051,74)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(47,24)     (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(3051,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(47,123)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(47,123)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(47,123)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,7)
│
├── train03/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(6537,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(6537,74)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(103,24)    (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(6537,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(103,144)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(103,144)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(103,144)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,15)
│
├── test03/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(3034,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(3034,76)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(47,24)     (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(3034,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(47,127)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(47,127)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(47,127)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,7)
│
├── train04/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(6520,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(6520,74)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(103,24)    (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(6520,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(103,127)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(103,127)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(103,127)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,15)
│
├── test04/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(3060,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(3060,73)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(47,24)     (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(3060,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(47,144)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(47,144)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(47,144)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,7)
│
├── train05/
│   ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
│   ├── boxes             # dtype=np.int32, shape=(6542,4)
│   ├── image_filenames   # dtype=np.uint8, shape=(6542,76)   (note: string in ASCII format)
│   ├── videos            # dtype=np.uint8, shape=(103,24)    (note: string in ASCII format)
│   ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
│   ├── object_ids        # dtype=np.int32, shape=(6542,4)
│   ├── list_boxes_per_video        # dtype=np.int32, shape=(103,144)
│   ├── list_filenames_per_video    # dtype=np.int32, shape=(103,144)
│   ├── list_object_ids_per_video   # dtype=np.int32, shape=(103,144)
│   └── list_videos_per_activity    # dtype=np.int32, shape=(10,15)
│
└── test05/
    ├── activities        # dtype=np.uint8, shape=(10,14)     (note: string in ASCII format)
    ├── boxes             # dtype=np.int32, shape=(3038,4)
    ├── image_filenames   # dtype=np.uint8, shape=(3038,75)   (note: string in ASCII format)
    ├── videos            # dtype=np.uint8, shape=(47,24)     (note: string in ASCII format)
    ├── object_fields     # dtype=np.uint8, shape=(4,16)      (note: string in ASCII format)
    ├── object_ids        # dtype=np.int32, shape=(3038,4)
    ├── list_boxes_per_video        # dtype=np.int32, shape=(47,127)
    ├── list_filenames_per_video    # dtype=np.int32, shape=(47,127)
    ├── list_object_ids_per_video   # dtype=np.int32, shape=(47,127)
    └── list_videos_per_activity    # dtype=np.int32, shape=(10,7)

Fields

  • activities: activity names
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.uint8
    • is padded: True
    • fill value: 0
    • note: strings stored in ASCII format
  • image_filenames: image file path+name
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.uint8
    • is padded: True
    • fill value: 0
    • note: strings stored in ASCII format
  • boxes: bounding box coordinates
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.int32
    • is padded: False
    • fill value: -1
    • note: bbox format [x1,y1,x2,y2]
  • videos: video name
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.uint8
    • is padded: True
    • fill value: 0
    • note: strings stored in ASCII format
  • object_fields: list of field names of the object id list
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.uint8
    • is padded: True
    • fill value: 0
    • note: strings stored in ASCII format
    • note: key field (field name aggregator)
  • object_ids: list of field ids
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.int32
    • is padded: False
    • fill value: -1
    • note: key field (field id aggregator)
  • list_boxes_per_video: list of bounding box ids per video
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.int32
    • is padded: True
    • fill value: -1
    • note: pre-ordered list
  • list_filenames_per_video: list of image ids per video
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.int32
    • is padded: True
    • fill value: -1
    • note: pre-ordered list
  • list_object_ids_per_video: list of object ids per video
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.int32
    • is padded: True
    • fill value: -1
    • note: pre-ordered list
  • list_videos_per_activity: list of video ids per activity
    • available in: train01, train02, train03, train04, train05, test01, test02, test03, test04, test05
    • dtype: np.int32
    • is padded: True
    • fill value: -1
    • note: pre-ordered list

Disclaimer

All rights reserved to the original creators of UCF-Sports.

For information about the dataset and its terms of use, please see this link.