A Modified Sparse Representation Method for Facial Expression Recognition

Figure 10

Images after being added to noise (variance from left to right is 0, 0.01, 0.05, 0.1, 0.2, 0.3, 0.4, and 0.5).