Commit 2bbecc51 authored by Davis King's avatar Davis King

updated docs

parent 5b3859a1
......@@ -126,6 +126,7 @@
<item>fine_hog_image</item>
<item>poly_image</item>
<item>hashed_feature_image</item>
<item>binned_vector_feature_image</item>
<item>nearest_neighbor_feature_image</item>
<item>randomly_sample_image_features</item>
</section>
......@@ -430,6 +431,40 @@
</component>
<!-- ************************************************************************* -->
<component>
<name>binned_vector_feature_image</name>
<file>dlib/image_keypoint.h</file>
<spec_file link="true">dlib/image_keypoint/binned_vector_feature_image_abstract.h</spec_file>
<description>
This object is a tool for performing image feature extraction. In
particular, it wraps another image feature extractor and converts the
wrapped image feature vectors into a high dimensional sparse vector. For
example, if the lower level feature extractor outputs the vector [3,4,5]
and this vector is hashed into the second bin of four bins then the output
sparse vector is:
<blockquote>
[0,0,0,0, 3,4,5,1, 0,0,0,0, 0,0,0,0].
</blockquote>
That is, the output vector has a dimensionality that is equal to the number
of hash bins times the dimensionality of the lower level vector plus one.
The value in the extra dimension concatenated onto the end of the vector is
always a constant value of of 1 and serves as a bias value. This means
that, if there are N hash bins, these vectors are capable of representing N
different linear functions, each operating on the vectors that fall into
their corresponding hash bin.
<br/><br/>
The following feature extractors can be wrapped by the binned_vector_feature_image:
<ul style="margin-top:0em">
<li><a href="#hog_image">hog_image</a></li>
<li><a href="#fine_hog_image">fine_hog_image</a></li>
<li><a href="#poly_image">poly_image</a></li>
</ul>
</description>
</component>
<!-- ************************************************************************* -->
<component>
......@@ -1623,6 +1658,7 @@
The following feature extractors can be used with the scan_image_pyramid object:
<ul style="margin-top:0em">
<li><a href="#hashed_feature_image">hashed_feature_image</a></li>
<li><a href="#binned_vector_feature_image">binned_vector_feature_image</a></li>
<li><a href="#nearest_neighbor_feature_image">nearest_neighbor_feature_image</a></li>
</ul>
</description>
......@@ -1666,6 +1702,7 @@
The following feature extractors can be used with the scan_image_boxes object:
<ul style="margin-top:0em">
<li><a href="#hashed_feature_image">hashed_feature_image</a></li>
<li><a href="#binned_vector_feature_image">binned_vector_feature_image</a></li>
<li><a href="#nearest_neighbor_feature_image">nearest_neighbor_feature_image</a></li>
</ul>
</description>
......@@ -1844,7 +1881,8 @@
<file>dlib/image_processing.h</file>
<spec_file link="true">dlib/image_processing/setup_hashed_features_abstract.h</spec_file>
<description>
This is a tool for configuring the <a href="#hashed_feature_image">hashed_feature_image</a> object
This is a tool for configuring the <a href="#hashed_feature_image">hashed_feature_image</a>
or <a href="#binned_vector_feature_image">binned_vector_feature_image</a> object
with a random <a href="algorithms.html#projection_hash">projection hash</a>.
</description>
<examples>
......
......@@ -1204,6 +1204,7 @@
<term file="imaging.html" name="fine_hog_image"/>
<term file="imaging.html" name="poly_image"/>
<term file="imaging.html" name="hashed_feature_image"/>
<term file="imaging.html" name="binned_vector_feature_image"/>
<term file="imaging.html" name="nearest_neighbor_feature_image"/>
<term file="imaging.html" name="haar_x"/>
<term file="imaging.html" name="haar_y"/>
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment