NextCloud/dlib
1
0
Fork 0
mirror of https://github.com/davisking/dlib.git synced 2024-11-01 10:14:53 +08:00
Code Issues Packages Projects Releases Wiki Activity

Clarified the spec regarding the type of sparse vector supported

Browse Source 
by these two objects.
This commit is contained in:
Davis King 2011-10-30 09:22:41 -04:00
parent 2cce156d84
commit 2116cbb763
2 changed files with 2 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
dlib/image_keypoint/hashed_feature_image_abstract.h
View File

@ -199,15 +199,12 @@ namespace dlib
- 0 <= col < nc()
ensures
- hashes BASE_FE(row,col) and returns the resulting indicator vector.
This vector will be represented as an unsorted sparse vector.
- Returns a vector V such that:
- V.size() == get_hash_bin_sizes().size()
- for all valid i: 0 <= V[i].first < get_num_dimensions()
- if (BASE_FE(row,col) hashes into bin B) then
- V contains an element with .first == B and .second == 1
- Note that the returned vector is represented as a sparse vector but
the indices are not in sorted order. Moreover, there might even be
duplicate entires for a particular dimension. This means you can't use
many of the sparse vector operations defined in dlib::sparse_vector.
!*/
const rectangle get_block_rect (

2
dlib/svm/structural_svm_problem_abstract.h
View File

@ -25,7 +25,7 @@ namespace dlib
REQUIREMENTS ON feature_vector_type_
- feature_vector_type_ == a dlib::matrix capable of storing column vectors
or a sparse vector type as defined in dlib/svm/sparse_vector_abstract.h.
or an unsorted sparse vector type as defined in dlib/svm/sparse_vector_abstract.h.
INITIAL VALUE
- get_epsilon() == 0.001