The DynTex database Homepage

The DynTex database

Home
Context
Protocol
Database
Classification  benchmark Annotations
Registration
Download
References

Classification Benchmark


In order to use DynTex as a benchmark for recognition, different sub-datasets have been  compiled and labeled. Only a single dynamic texture is present in each sequence with no  panning or zooming performed by the camera.

The DynTex team thanks Sloven Dubois for compiling thoses datasets

 Alpha dataset

The Alpha dataset is composed of 60 dynamic textures divided into 3 classes.
The list of the concerned video files is linked here.
  • Sea
    Image alpha_1
  • Grass 
    Image alpha_2
  • Trees
    Image alpha_3



Beta dataset

The Beta dataset is composed of 162 dynamic textures divided into 10 classes.
The list of the concerned video files is linked here.
  • Sea
    Image beta_1
  • Vegetation
    Image beta_2
  • Trees
    Image beta_3
  • Flags
    Image beta_4
  • Calm water
    Image beta_5
  • Fountains
    Image beta_6
  • Smoke
    Image beta_7
  • Escalator
    Image beta_8
  • Traffic
    Image beta_9
  • Rotation
    Image beta_10



 Gamma Dataset

The Gamma dataset is composed of 275 dynamic textures divided into 10 classes.
The list of the concerned video files is linked here.
  • Flowers
    Image gamma_1
  • Sea
    Image gamma_2
  • Naked trees
    Image gamma_3
  • Foliage
    Image gamma_4
  • Escalator
    Image gamma_5
  • Calm water
    Image gamma_6
  • Flags
    Image gamma_7
  • Grass
    Image gamma_8
  • Traffic
    Image gamma_9
  • Fountains
    Image gamma_10



Remarks? Questions? Broken links?  mail!mail_dyntex