Skip to content

Commit

Permalink
Update docs
Browse files Browse the repository at this point in the history
  • Loading branch information
mattdawkins committed Feb 5, 2018
1 parent 2db706c commit e8a3d4d
Showing 1 changed file with 16 additions and 15 deletions.
31 changes: 16 additions & 15 deletions examples/image_and_video_search/README.rst
Original file line number Diff line number Diff line change
@@ -1,8 +1,22 @@

======================
Rapid Model Generation
======================

Rapid model generation can be performed using the same method as image and video search
(below), just saving out the resultant trained detection models after performing iterative
query refinement.

Additional documentation coming soon.

===============================
Image and Video Search Examples
===============================

********
Overview
********

This directory contains methods to accomplish two tasks:

|
Expand All @@ -12,9 +26,8 @@ This directory contains methods to accomplish two tasks:
Only Linux and Mac are currently supported for running these examples.

********************
Image Archive Search
********************
====================

The "smqtk_on_chips" directory contains multiple methods for running image
queries on an image archive, including:
Expand All @@ -32,9 +45,8 @@ IQR generates an SVM classifier on user-nominated positive and negative
examples from the result set to refine results.


********************
Video Archive Search
********************
====================

Video archive search can be performed via a few methods. The default includes
a pipeline which generates object detections, tracks, and lastly temporal
Expand All @@ -44,14 +56,3 @@ table, or other). At query time, descriptors on a query image or video are
matched against the entries in this database. A default GUI (provided via
the VIVIA toolkit) is provided which has similar capabilities to the SMQTK
GUI, ie, search and IQR.


======================
Rapid Model Generation
======================

Rapid model generation can be performed in the same way as image and video search,
just saving out the resultant trained detection models after performing iterative
query refinement.

Additional documentation coming soon.

0 comments on commit e8a3d4d

Please sign in to comment.