Skip to content

Releases: matlab-deep-learning/mtcnn-face-detection

MTCNN Face Detection v1.2.4

03 Mar 20:54
503f4ac
Compare
Choose a tag to compare
  • Fixes bug when using a GPU in R2020b (#20)

MTCNN Face Detection v1.2.3

14 Sep 14:43
3ec4d03
Compare
Choose a tag to compare
  • Fixes bug in relative ordering of boxes and landmarks as well as potential error when there are overlapping boxes (#17)
  • 2 point version increase to be able to use GitHub releases integration on File Exchange.

MTCNN Face Detection v1.2.1

12 Apr 20:03
a0a8ecc
Compare
Choose a tag to compare
  • Fixes bug when using UseGPU=True with UseDagNet=True (i.e. R2019a support) (#12)

MTCNN Face Detection v1.2.0

01 Apr 13:19
fb1fec4
Compare
Choose a tag to compare
  • Adds support for R2019a
    • Uses DAGNetwork versions of the CNNs for compatibility with R2019a
  • detect and detectFaces functions accept uint8, single, and double image types
    • singles and doubles are expected to be in the range 0-1
    • uint8s are expected to be in the range 0-255
  • fixes a bug in proposeRegions for certain image sizes.

MTCNN Face Detection v1.0.1

22 Jan 08:43
8bfcbca
Compare
Choose a tag to compare
  • Adds support for Name Value pair optional arguments to mtcnn.detectFaces

MTCNN Face Detection v1.0

16 Jan 08:44
65364d6
Compare
Choose a tag to compare

Version 1.0 allows face detection and alignment using a pre-trained MTCNN detector.

To install download the .mltbx file and open using MATLAB.

Requirements:

  • MATLAB R2019b or later
  • Deep Learning Toolbox
  • Computer Vision Toolbox
  • Image Processing Toolbox