Datasets

Dog Detection

Dog Detection

The dataset features 1498 images of dogs in various poses which are encountered during the daily commute in an Indian context. The data has been captured through a wearable POV Camera (Pivothead). Annotations provide:

  • Bounding Box
  • Pose
  • Download here
    Signboard Detection

    SignBoard Detection

    1493 images for 85 signboard. Annotations contain:

  • Bounding Box
  • Text
  • Direction
  • Type of signboard
  • Illumination information
  • Download here
    Texture Classification

    Texture Classification

    An important insight for navigation for a visually impaired person is the kind of textures they are going to step into. In this regard, we release the dataset of texture classification. Annotations contain the classification of blocks in a rectangular grid as Pavement, Road, Grass or Other.
    Download here

    Animal Detection

    Cow Detection

    Consists of 1604 images. Following annotations are provided:

  • Bounding Box
  • Color
  • Pose
  • Occluded/Not Occluded
  • Download here

    If this dataset has proved useful for you in your research, kindly cite us:

    Rajesh Kedia, Anupam Sobti, Mukund Rungta, Sarvesh Chandoliya, Akhil Soni, Anil Kumar Meena, Chrystle Myrna Lobo, Richa Verma, M. Balakrishnan, and Chetan Arora. 2019. MAVI: Mobility assistant for visually impaired with optional use of local and cloud resources. Accepted for publication. In 2019 32nd International Conference on VLSI Design and 2019 18th International Conference on Embedded Systems (VLSID).

    Bibtex:

    @INPROCEEDINGS{Kedia:2019:MAVI,
    author={Rajesh Kedia and Anupam Sobti and Mukund Rungta and Sarvesh Chandoliya and Akhil Soni and Anil Kumar Meena and Chrystle Myrna Lobo and Richa Verma and M. Balakrishnan and Chetan Arora},
    booktitle={2019 32nd International Conference on VLSI Design and 2019 18th International Conference on Embedded Systems (VLSID)},
    title={MAVI: Mobility assistant for visually impaired with optional use of local and cloud resources},
    year={2019},
    month={Jan},}