1214 |
} |
} |
1215 |
\end{figure} |
\end{figure} |
1216 |
|
|
1217 |
\subsection{Recognizability and memorizability} |
\subsection{Recognizability experiment} |
1218 |
|
|
1219 |
%In our application visual texture discrimination is not as |
%In our application visual texture discrimination is not as |
1220 |
%much of an issue as memorizability and recognizability of |
%much of an issue as memorizability and recognizability of |
1227 |
|
|
1228 |
Experiments on black-and-white %(faces,) |
Experiments on black-and-white %(faces,) |
1229 |
ink blots and snow crystals |
ink blots and snow crystals |
1230 |
\cite{goldstein71visualrecognition} show that |
\cite{goldstein71visualrecognition} have shown that |
1231 |
complex, unfamiliar pictures can be remembered and recognized |
complex, unfamiliar pictures can be remembered and recognized |
1232 |
and that recognition |
and that recognition |
1233 |
performance decreases very little over time. |
performance decreases very little over time. |
1234 |
|
|
1235 |
In order to evaluate the recognizability of our procedurally generated |
In order to evaluate the recognizability of our procedurally generated |
1236 |
textures, |
textures, we need to have an appropriate comparison point. |
1237 |
|
% |
1238 |
|
%- need to compare to something |
1239 |
|
%- can't generate an infinite supply |
1240 |
|
% of images easily and there are too much semantics |
1241 |
|
Pictures of natural objects would not provide a good comparsion, |
1242 |
|
because they cannot be generated in infinite amounts from seed |
1243 |
|
numbers and they would esasily yield undesirable semantic associations. |
1244 |
|
Thus, we simply use plain solid color backgrounds as a baseline for comparison |
1245 |
|
even though the colors of even a small set of randomly chosen colors would |
1246 |
|
most likely not be discriminable. |
1247 |
|
|
1248 |
- need to compare to something |
%even though know solid colors not really recognizable. |
1249 |
|
|
1250 |
- images not good, can't generate infinite supply easily, too much |
Surely it would be hard to remember a large number of textures, too. |
|
semantics |
|
|
|
|
|
- thus, use solid colors even though know solid colors not really recognizable. |
|
|
|
|
|
|
|
|
Surely it would be hard to remember, e.g., 100 textures. |
|
1251 |
However, the user does not have to remember all the textures; |
However, the user does not have to remember all the textures; |
1252 |
it suffices to learn the textures of the most often used documents |
it suffices to learn the textures of the most often used documents |
1253 |
and distinguish them from the rest: |
and distinguish them from the rest: |
1271 |
Each square represents a document, and the area of each square is scaled |
Each square represents a document, and the area of each square is scaled |
1272 |
to its rate of accesses. |
to its rate of accesses. |
1273 |
The diagram shows 2000 documents weighted with Zipf's law with exponent 1.1. |
The diagram shows 2000 documents weighted with Zipf's law with exponent 1.1. |
1274 |
The 15 most important documents account XXX\% of the accesses. |
The 15 most important documents account for 50\% of the accesses. |
1275 |
|
% 0.50469672124463749 |
1276 |
} |
} |
1277 |
\end{figure} |
\end{figure} |
1278 |
|
|
1279 |
Thus, the experiment measured the recognition of only 15 textures, |
Thus, we chose to measure the recognition of only 15 target textures |
1280 |
and |
or colors in the experiment. |
1281 |
|
Our hypthesis is that the texture backgrounds are more recognizable |
1282 |
|
than the solid colors. |
1283 |
|
|
1284 |
Furthermore, because the texture appearance has no correlation |
%and |
1285 |
|
% |
1286 |
|
|
1287 |
|
XXX: Furthermore, because the texture appearance has no correlation |
1288 |
with the document content, |
with the document content, |
1289 |
the textures of any two important files are similar only by chance. |
the textures of any two important files are similar only by chance. |
1290 |
|
|
1291 |
We have conducted an experiment in a similar setting comparing |
%We have conducted an experiment in a similar setting comparing |
1292 |
the recognition performance of our textures with solid color backgrounds. |
%the recognition performance of our textures with solid color backgrounds. |
|
|
|
1293 |
|
|
1294 |
|
\subsubsection{Method} |
1295 |
|
|
1296 |
Five participants performed the experiment in both conditions |
\emph{Participants.} |
1297 |
(in random order). |
Five participants naïve to the hypothesis and our texturing work |
1298 |
|
performed the experiment in both conditions. |
1299 |
|
|
1300 |
|
\emph{Design and materials.} |
1301 |
15 target backgrounds and 15 distractor backgrounds were randomly |
15 target backgrounds and 15 distractor backgrounds were randomly |
1302 |
chosen for both conditions. |
chosen for both conditions. |
1303 |
The distribution of the solid colors was the same that is used for |
The distribution of the solid colors was the same that is used for |
1305 |
lighness tail was de-emphasized to increase the otherwise low |
lighness tail was de-emphasized to increase the otherwise low |
1306 |
discriminability of very light, unsaturated colors. |
discriminability of very light, unsaturated colors. |
1307 |
|
|
1308 |
|
\emph{Procedure.} |
1309 |
|
Each participant performed the test individually in both conditions |
1310 |
|
in a random order. |
1311 |
First, the 15 target backgrounds were shown sequentially, |
First, the 15 target backgrounds were shown sequentially, |
1312 |
in a random order, 5 seconds each. |
in a random order, 5 seconds each. |
1313 |
Then, recognition was tested by showing the 15 target backgrounds |
Then, recognition was tested by showing the 15 target backgrounds |
1314 |
and 15 distractor backgrounds in a random order and |
and the 15 distractor backgrounds in a random order and |
1315 |
having the participant answer ``old'' or ``new'' for each one. |
having the participant answer ``old'' or ``new'' for each one. |
1316 |
The time for answers was not limited. |
The time for answers was not limited. |
1317 |
|
|
1318 |
|
\subsubsection{Results} |
1319 |
|
|
1320 |
The results are summarized in Table~\ref{tableresults} |
The results are summarized in Table~\ref{tableresults} |
1321 |
|
|
1322 |
\begin{table} |
\begin{table} |
1347 |
%use mutual information or something like that. In colors, the subject |
%use mutual information or something like that. In colors, the subject |
1348 |
%was basically guessing!!! |
%was basically guessing!!! |
1349 |
|
|
|
% RT less for correct responses: [F(1,4) = 5.2, p = .082] |
|
1350 |
Thus, the recognition performance of the textures was good while |
Thus, the recognition performance of the textures was good while |
1351 |
the solid colors do not have enough variation for unambiguous |
the solid colors do not have enough variation for unambiguous |
1352 |
recognition (without additional cues). |
recognition. |
1353 |
The observed difference in recognition performance is statistically |
A repeated measures analysis of variance indicates that |
1354 |
|
the observed difference in the recognition performance is statistically |
1355 |
significant [F(1,4) = 19.0, p = .012]. |
significant [F(1,4) = 19.0, p = .012]. |
1356 |
|
% RT less for correct responses: [F(1,4) = 5.2, p = .082] |
1357 |
|
|
1358 |
%Our experience shows that at least the most recurring textures |
%Our experience shows that at least the most recurring textures |
1359 |
%are not only recognized but can |
%are not only recognized but can |
1360 |
%also easily be associated with the document content. |
%also easily be associated with the document content. |
1361 |
|
|
1362 |
|
The solid colors could be more recognizable with additional cues, |
1363 |
|
such as the text of the document, but |
1364 |
|
|
1365 |
|
- fragments and not title page etc... |
1366 |
|
|
1367 |
XXX: refs? |
XXX: refs? |
1368 |
|
|
1369 |
%\section{Software availability} |
%\section{Software availability} |