Related Publications

This set of data is used in the following publications:


There are three sites on the USC campus:

These sites are selected because they present a different challange visually: a human made building complex, a park full of trees, and an open field area. The sites can be viewed in the following map (figure 1).

Figure 1. Map of testing sites

The data is taken using a camera. The videos are arranged as follows: each column lists videos from a particular path segment (there are 9 for every site, observe the site map for the segment locations). Each row is designated for a particular trial run (which means that a row has the same lighting condition), except for ACB. This is because ACB is the first testing site and, at that time, we were still trying to figure out the best way to train a neural network that would have lighting invariance.

Selection of filming times that include all lighting conditions is the main issue in collecting training samples. Because lighting space is hard to gauge, it takes several days to come up with times of day to cover most of them. At the end settle on the following times: early morning, noon, early afternoon, late afternoon, and early evening. We do not include the night time videos because some sites are simply too dark.

For testing episodes, we tape two dark and two bright conditions separately (they are not used for training). Also note that there are also other significant discrepancies in forms of people walking, missing objects (a park bench taken away), and new objects (service vehicles) between the training and testing dataset.

Ahmanson Center Of Biological Science (ACB)

The ACB scenes (shown in figure 2) are filmed throughout the narrow corridors of the building complex. Most of the surroundings are flat walls with little texture and solid lines that delineate the walls and different parts of the buildings.

Figure 2. Map and example images of the ACB site

Figure 3 shows the different lighting conditions tested, from left to right: late afternoon (trial 1), early evening with lights already turned on (trial 2), mid-afternoon (trial 3), and noon (trial 4).

Figure 3. Lighting conditions for the ACB site testing

Run/Segment Number 1 2 3 4 5 6 7 8 9
A ACB1A.mpg ACB2A.mpg ACB3A.mpg ACB4A.mpg ACB5A.mpg ACB6A.mpg ACB7A.mpg ACB8A.mpg ACB9A.mpg
B ACB1B.mpg ACB2B.mpg ACB3B.mpg ACB4B.mpg ACB5B.mpg ACB6B.mpg ACB7B.mpg ACB8B.mpg ACB9B.mpg
C ACB1C.mpg ACB2C.mpg ACB3C.mpg ACB4C.mpg ACB5C.mpg ACB6C.mpg ACB7C.mpg ACB8C.mpg ACB9C.mpg
D ACB1D.mpg ACB2D.mpg ACB3D.mpg ACB4D.mpg ACB5D.mpg ACB6D.mpg ACB7D.mpg ACB8D.mpg ACB9D.mpg
E ACB1E.mpg ACB2E.mpg ACB3E.mpg ACB4E.mpg ACB5E.mpg ACB6E.mpg ACB7E.mpg ACB8E.mpg ACB9E.mpg
F ACB1F.mpg ACB2F.mpg ACB3F.mpg ACB4F.mpg ACB5F.mpg ACB6F.mpg ACB7F.mpg ACB8F.mpg ACB9F.mpg
G ACB1G.mpg   ACB3G.mpg ACB4G.mpg ACB5G.mpg     ACB8G.mpg  
H ACB1H.mpg                

Run/Segment Number 1 2 3 4 5 6 7 8 9
A ACB_T_1A.mpg ACB_T_2A.mpg ACB_T_3A.mpg ACB_T_4A.mpg ACB_T_5A.mpg ACB_T_6A.mpg ACB_T_7A.mpg ACB_T_8A.mpg ACB_T_9A.mpg
B ACB_T_1B.mpg ACB_T_2B.mpg ACB_T_3B.mpg ACB_T_4B.mpg ACB_T_5B.mpg ACB_T_6B.mpg ACB_T_7B.mpg ACB_T_8B.mpg ACB_T_9B.mpg
C ACB_T_1C.mpg ACB_T_2C.mpg ACB_T_3C.mpg ACB_T_4C.mpg ACB_T_5C.mpg ACB_T_6C.mpg ACB_T_7C.mpg ACB_T_8C.mpg ACB_T_9C.mpg
D ACB_T_1D.mpg ACB_T_2D.mpg ACB_T_3D.mpg ACB_T_4D.mpg ACB_T_5D.mpg ACB_T_6D.mpg ACB_T_7D.mpg ACB_T_8D.mpg ACB_T_9D.mpg
E ACB_T_1E.mpg ACB_T_2E.mpg ACB_T_3E.mpg ACB_T_4E.mpg ACB_T_5E.mpg ACB_T_6E.mpg ACB_T_7E.mpg ACB_T_8E.mpg ACB_T_9E.mpg
F ACB_T_1F.mpg ACB_T_2F.mpg ACB_T_3F.mpg ACB_T_4F.mpg ACB_T_5F.mpg ACB_T_6F.mpg ACB_T_7F.mpg ACB_T_8F.mpg ACB_T_9F.mpg

The entire ACB data in one .tar file: ACB.tar.gz

Associate and Founder's Park (AnF)

The scenes of AnF (figure 4), are dominated by vegetations. There are fewer extractable structures as large areas of the images are practically un-segmentable as leaves overrun most regions. In addition, the length of the segments at this site are about twice the lengths of the segments in ACB.

Figure 4. Map and example images of the AnF site

Figure 5 shows the different lighting conditions tested, from left to right: overcast (trial 1), early evening with lights already turned on (trial 2), mid-afternoon (trial 3), and noon (trial 4).

Figure 5. Lighting conditions for the AnF site testing

Run/Segment Number 1 2 3 4 5 6 7 8 9
A AnFpark1A.mpg AnFpark2A.mpg AnFpark3A.mpg AnFpark4A.mpg AnFpark5A.mpg AnFpark6A.mpg AnFpark7A.mpg AnFpark8A.mpg AnFpark9A.mpg
B AnFpark1B.mpg AnFpark2B.mpg AnFpark3B.mpg AnFpark4B.mpg AnFpark5B.mpg AnFpark6B.mpg AnFpark7B.mpg AnFpark8B.mpg AnFpark9B.mpg
C AnFpark1C.mpg AnFpark2C.mpg AnFpark3C.mpg AnFpark4C.mpg AnFpark5C.mpg AnFpark6C.mpg AnFpark7C.mpg AnFpark8C.mpg AnFpark9C.mpg
D AnFpark1D.mpg AnFpark2D.mpg AnFpark3D.mpg AnFpark4D.mpg AnFpark5D.mpg AnFpark6D.mpg AnFpark7D.mpg AnFpark8D.mpg AnFpark9D.mpg
E AnFpark1E.mpg AnFpark2E.mpg AnFpark3E.mpg AnFpark4E.mpg AnFpark5E.mpg AnFpark6E.mpg AnFpark7E.mpg AnFpark8E.mpg AnFpark9E.mpg
F AnFpark1F.mpg AnFpark2F.mpg AnFpark3F.mpg AnFpark4F.mpg AnFpark5F.mpg AnFpark6F.mpg AnFpark7F.mpg AnFpark8F.mpg AnFpark9F.mpg
G AnFpark1G.mpg AnFpark2G.mpg AnFpark3G.mpg AnFpark4G.mpg AnFpark5G.mpg AnFpark6G.mpg AnFpark7G.mpg AnFpark8G.mpg AnFpark9G.mpg
H AnFpark1H.mpg AnFpark2H.mpg AnFpark3H.mpg AnFpark4H.mpg AnFpark5H.mpg AnFpark6H.mpg AnFpark7H.mpg AnFpark8H.mpg AnFpark9H.mpg
I AnFpark1I.mpg AnFpark2I.mpg AnFpark3I.mpg AnFpark4I.mpg AnFpark5I.mpg AnFpark6I.mpg AnFpark7I.mpg AnFpark8I.mpg AnFpark9I.mpg
J AnFpark1J.mpg AnFpark2J.mpg AnFpark3J.mpg AnFpark4J.mpg AnFpark5J.mpg AnFpark6J.mpg AnFpark7J.mpg AnFpark8J.mpg AnFpark9J.mpg

Run/Segment Number 1 2 3 4 5 6 7 8 9
A AnFpark_T_1A.mpg AnFpark_T_2A.mpg AnFpark_T_3A.mpg AnFpark_T_4A.mpg AnFpark_T_5A.mpg AnFpark_T_6A.mpg AnFpark_T_7A.mpg AnFpark_T_8A.mpg AnFpark_T_9A.mpg
B AnFpark_T_1B.mpg AnFpark_T_2B.mpg AnFpark_T_3B.mpg AnFpark_T_4B.mpg AnFpark_T_5B.mpg AnFpark_T_6B.mpg AnFpark_T_7B.mpg AnFpark_T_8B.mpg AnFpark_T_9B.mpg
C AnFpark_T_1C.mpg AnFpark_T_2C.mpg AnFpark_T_3C.mpg AnFpark_T_4C.mpg AnFpark_T_5C.mpg AnFpark_T_6C.mpg AnFpark_T_7C.mpg AnFpark_T_8C.mpg AnFpark_T_9C.mpg
D AnFpark_T_1D.mpg AnFpark_T_2D.mpg AnFpark_T_3D.mpg AnFpark_T_4D.mpg AnFpark_T_5D.mpg AnFpark_T_6D.mpg AnFpark_T_7D.mpg AnFpark_T_8D.mpg AnFpark_T_9D.mpg

The entire AnFpark data in one .tar file: AnFpark.tar.gz

Frederick D. Fagg (FDF)

A large portion of the FDF site (figure 6) is sky, mostly textureless space with scattered light clouds. The main reason for testing at this site is to assess the gist response on sparser data. The length of the segments are about 50 percent longer than the ones in the second experiment (three times that of ACB).

Figure 6. Map and example images of the FDF site

Figure 7 shows the different lighting conditions tested,from left to right: early evening (street lights are off), early evening (street lights are already on), noon, and middle of afternoon, with the dramatic long shadow cast on the field by a tall building.

Figure 7. Lighting conditions for the FDF site testing

Run/Segment Number 1 2 3 4 5 6 7 8 9
A FDFpark1A.mpg FDFpark2A.mpg FDFpark3A.mpg FDFpark4A.mpg FDFpark5A.mpg FDFpark6A.mpg FDFpark7A.mpg FDFpark8A.mpg FDFpark9A.mpg
B FDFpark1B.mpg FDFpark2B.mpg FDFpark3B.mpg FDFpark4B.mpg FDFpark5B.mpg FDFpark6B.mpg FDFpark7B.mpg FDFpark8B.mpg FDFpark9B.mpg
C FDFpark1C.mpg FDFpark2C.mpg FDFpark3C.mpg FDFpark4C.mpg FDFpark5C.mpg FDFpark6C.mpg FDFpark7C.mpg FDFpark8C.mpg FDFpark9C.mpg
D FDFpark1D.mpg FDFpark2D.mpg FDFpark3D.mpg FDFpark4D.mpg FDFpark5D.mpg FDFpark6D.mpg FDFpark7D.mpg FDFpark8D.mpg FDFpark9D.mpg
E FDFpark1E.mpg FDFpark2E.mpg FDFpark3E.mpg FDFpark4E.mpg FDFpark5E.mpg FDFpark6E.mpg FDFpark7E.mpg FDFpark8E.mpg FDFpark9E.mpg
F FDFpark1F.mpg FDFpark2F.mpg FDFpark3F.mpg FDFpark4F.mpg FDFpark5F.mpg FDFpark6F.mpg FDFpark7F.mpg FDFpark8F.mpg FDFpark9F.mpg
G FDFpark1G.mpg FDFpark2G.mpg FDFpark3G.mpg FDFpark4G.mpg FDFpark5G.mpg FDFpark6G.mpg FDFpark7G.mpg FDFpark8G.mpg FDFpark9G.mpg

Run/Segment Number 1 2 3 4 5 6 7 8 9
A FDFpark_T_1A.mpg FDFpark_T_2A.mpg FDFpark_T_3A.mpg FDFpark_T_4A.mpg FDFpark_T_5A.mpg FDFpark_T_6A.mpg FDFpark_T_7A.mpg FDFpark_T_8A.mpg FDFpark_T_9A.mpg
B FDFpark_T_1B.mpg FDFpark_T_2B.mpg FDFpark_T_3B.mpg FDFpark_T_4B.mpg FDFpark_T_5B.mpg FDFpark_T_6B.mpg FDFpark_T_7B.mpg FDFpark_T_8B.mpg FDFpark_T_9B.mpg
C FDFpark_T_1C.mpg FDFpark_T_2C.mpg FDFpark_T_3C.mpg FDFpark_T_4C.mpg FDFpark_T_5C.mpg FDFpark_T_6C.mpg FDFpark_T_7C.mpg FDFpark_T_8C.mpg FDFpark_T_9C.mpg
D FDFpark_T_1D.mpg FDFpark_T_2D.mpg FDFpark_T_3D.mpg FDFpark_T_4D.mpg FDFpark_T_5D.mpg FDFpark_T_6D.mpg FDFpark_T_7D.mpg FDFpark_T_8D.mpg FDFpark_T_9D.mpg
H FDFpark_T_1H.mpg FDFpark_T_2H.mpg FDFpark_T_3H.mpg FDFpark_T_4H.mpg FDFpark_T_5H.mpg FDFpark_T_6H.mpg FDFpark_T_7H.mpg FDFpark_T_8H.mpg FDFpark_T_9H.mpg
I FDFpark_T_1I.mpg FDFpark_T_2I.mpg FDFpark_T_3I.mpg FDFpark_T_4I.mpg FDFpark_T_5I.mpg FDFpark_T_6I.mpg FDFpark_T_7I.mpg FDFpark_T_8I.mpg FDFpark_T_9I.mpg
J FDFpark_T_1J.mpg FDFpark_T_2J.mpg FDFpark_T_3J.mpg FDFpark_T_4J.mpg FDFpark_T_5J.mpg FDFpark_T_6J.mpg FDFpark_T_7J.mpg FDFpark_T_8J.mpg FDFpark_T_9J.mpg
K FDFpark_T_1K.mpg FDFpark_T_2K.mpg FDFpark_T_3K.mpg FDFpark_T_4K.mpg FDFpark_T_5K.mpg FDFpark_T_6K.mpg FDFpark_T_7K.mpg FDFpark_T_8K.mpg FDFpark_T_9K.mpg

The entire FDFpark data in one .tar file: FDFpark.tar.gz

Copyright © 2000 by the University of Southern California, iLab and Prof. Laurent Itti