Presentation is loading. Please wait.

Presentation is loading. Please wait.

Computer examples Tenenbaum, de Silva, Langford “A Global Geometric Framework for Nonlinear Dimensionality Reduction”

Similar presentations


Presentation on theme: "Computer examples Tenenbaum, de Silva, Langford “A Global Geometric Framework for Nonlinear Dimensionality Reduction”"— Presentation transcript:

1 Computer examples Tenenbaum, de Silva, Langford “A Global Geometric Framework for Nonlinear Dimensionality Reduction”

2 Statue face database x64 grayscale images 2 mins, 12 secs on a ~600 (?) MHz PIII

3 The computed manifold

4

5 Testing the sensibility of the manifold coordinates One test you could do: 1.Sort all faces according to first manifold coordinate (“left-right”) 2.View them in order 3.See if the face makes a monotonic progression from left to right

6 Right  Left

7 Up  Down Cleaner, since light variation is strictly azimuthal (consistent chin shadow)

8 Lit on left  Lit on right

9 Testing the sensibility of the manifold coordinates Semantic consistency of a dimension value deteriorates between points that are far away on the manifold. 4 consecutive frames from right  left movie: Well-lit faces are turning to the left with respect to each other Dimly-lit faces also don’t turn right w.r.t each other

10 Testing the sensibility of the manifold coordinates Semantic consistency of a dimension value deteriorates between points that are far away on the manifold. Explanations: Geodesic distance on the manifold is approximated by shortest-path distance in a neighbor graph.  Sparsity in neighbor graphs result in distance error for points far away on the graph.

11 Testing the sensibility of the manifold coordinates Geodesic distance approximator can’t be perfect in the face of sparse data

12 Testing the sensibility of the manifold coordinates The test expected this face:

13 Testing the sensibility of the manifold coordinates …to be a bit more left-facing than this face:

14 Traversing the manifold Collapsing the manifold to one dimension isn’t the way to use it. Try tracing one dimension while keeping the other dimensions from jumping around too much.

15 Traversing the manifold Algorithm used: Sort images by “left-right” coord as before Draw a smooth line through the manifold Only add images that are within a certain manifold distance D from this line.

16 Traversing the manifold

17

18

19 D = 20 (Half the range of the “up-down” dimension)

20 Traversing the manifold (D = 30)

21 Traversing the manifold D = 40 (using 80% of the faces)

22 Traversing the manifold D = 50 (using 98% of the faces)

23 Comparison to LLE Run both algorithms on 100 of the statue faces (64 x 64 pixels) Isomap LLE

24 Comparison to LLE Running time for x64 images: LLE: 5 secs Isomap: 1.39 secs

25 Comparison to LLE The collapsing-to-primary-dimension-test:

26 Comparison to LLE Uh… the collapsing-to-second-dimension-test

27 Comparison to LLE The horizontal manifold traversal test (7 frames)

28 Comparison to LLE LLE: once manifold is computed, meaningful paths through it need to be searched for.

29 Weakness under translation Images with a common background and a single translating object will have a rough time with pixel differences.

30 Weakness under translation Uniform translation, no overlap Input images: Output images:

31 Weakness under translation Uniform translation, 1-column overlap Input images: Output images:

32 Weakness under translation Uniform translation, 1-column overlap

33 Weakness under translation Uniform translation, with a skip

34 Weakness under translation Isomap with k = 1 (like before) (Original) (Reconstruction)

35 Weakness under translation Isomap with k = 2 (Original) (Reconstruction)

36 Overestimating k Isomap with k = 2

37 End


Download ppt "Computer examples Tenenbaum, de Silva, Langford “A Global Geometric Framework for Nonlinear Dimensionality Reduction”"

Similar presentations


Ads by Google