TODO list for low level MRF models

[Gnimuc]

• [ ] 2 Mathematical MRF Models
  • [x] Ising Model
  • [x] Auto-Logistic Model
  • [ ] Auto-Binomial Model
  • [ ] Auto-Normal Model
  • [x] Multi-level Logistic Model (MLL)
• [ ] 3.2 Image Restoration and Reconstruction
• [x] 3.4 Texture Synthesis and Analysis
  • [x] Sampling Algorithm 1 (Metropolis)
  • [x] Sampling Algorithm 2 (Gibbs)
• [ ] 7.1 Supervised Estimation with Labeled Data
  • [x] Pseudo-likelihood
  • [ ] Least Squares Fit
  • [x] Markov Chain Monte Carlo Methods(MCMC)
  • [x] Metropolis-Hastings Algorithm
  • [ ] Gibbs Sampler
• [ ] 7.2 Unsupervised Estimation with Unlabeled Data
  • [ ] Simultaneous Restoration and Estimation
  • [ ] Simultaneous Segmentation and Estimation
• [ ] 9.3 Minimization with Discrete Labels
  • [ ] Iterated Conditional Modes
  • [ ] Belief Propagation
• [ ] 10.1 Simulated Annealing
• [ ] 10.3 Graduated Nonconvexity
• [ ] 10.4 Graph Cuts
  • [x] Max-Flow
  • [x] Multilabel Graph Cuts
  • [x] α-expansion
  • [ ] α–β Swap
• [ ] 10.5.2 Hybrid GA: Comb Algorithm
• [ ] 10.6 Experimental Comparisons

[Gnimuc]

@wudixiaopeng

[Gnimuc]

@wudixiaopeng 根据12年TIP那篇文章作者给的代码, 可以把9.3 ICM重写一下加进来. #9

[Gnimuc]

@501306162 建议挑几个label为up for grabs的issue实现一下，并review现有代码。

[Gnimuc]

不熟悉GitHub flow可以看看这个培训视频http://www.bilibili.com/video/av7893882/