Michel BARLAUD - research
Professor Michel BARLAUD

RESEARCH



  • Image and video coding
    • Wavelet Transform 2D (IEEE IP 1992).
    • Lattice Vector Quantization
    • t+2D Motion Compensated Temporal Lifting Transform
    • Optimal Decoding (IEEE MMA 2003).
    • Entropy-based distortion measure and bit allocation for wavelet image compression (IEEE IP 2007).
  • Inverse problems
    • Half Quadratic Regularization (IEEE IP 1997).
    • 3D Reconstruction of SPECT medical images
    • Deconvolution and image denoising
  • Segmentation and Moton estimationImage and video Analysis
    • Region Based Segmentation and Motion Estimation using Shape Gradient Method (IJCV 2003).
    • Segmentation uing shape Priors (IEEE CSVT 2003)
    • Parametric Active Contours Segmentation using "Smoothing Bsplines" (IEEE IP 2004)
    • Object tracking (IEEE CVPR 2007).
  • Machine learning for computer vision
    • Calibrated Loss for supervised classification IEEE PAMI
    • Stochastic Boosting algorithms
    • Convergence proofs using Bregman divergence

    • Supervised classification of images
    • Cyber security
    • Big data analysis
    • Biologic cells classifcation ICPR 2012