-
Notifications
You must be signed in to change notification settings - Fork 8
/
memorability.html
474 lines (405 loc) · 23.9 KB
/
memorability.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description" content="Long-Term Ad Memorability: Understanding and Generating Memorable Ads">
<meta name="keywords" content="memorability, advertisements, large language models">
<meta name="viewport" content="width=device-width, initial-scale=1">
<meta property="og:url" content="https://behavior-in-the-wild.github.io/memorability">
<title>Long-Term Ad Memorability</title>
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bulma@0.9.1/css/bulma.min.css">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/4.5.2/css/bootstrap.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.15.1/css/all.min.css">
<link rel="stylesheet" href="./static/css/base.css">
<link rel="icon" href="https://cdn-icons-png.flaticon.com/512/954/954591.png">
<link href="https://fonts.googleapis.com/icon?family=Material+Icons" rel="stylesheet">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script defer src="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.15.1/js/all.min.js"></script>
<script type="module" src="https://gradio.s3-us-west-2.amazonaws.com/4.16.0/gradio.js"></script>
<style>
body {
font-family: 'Noto Sans', sans-serif;
margin: 0;
padding: 0;
box-sizing: border-box;
}
header {
background-color: #333;
color: #fff;
padding: 10px 20px;
text-align: center;
}
.container {
padding: 20px;
}
.video-panel {
display: grid;
grid-template-columns: repeat(3, 1fr);
gap: 20px;
}
.video-item {
text-align: center;
}
.video-caption {
margin-top: 10px;
font-size: 18px;
}
.examples {
margin-top: 40px;
}
.example-item {
margin-bottom: 20px;
}
.memorable-trends-list {
font-size: 1.25em; /* Increase this value as needed */
line-height: 1.6;
}
.memorable-trends-list li {
margin-bottom: 10px;
}
</style>
</head>
<body>
<header class="header", style="background-color:#ff0202;">
<nav class="navbar", role="navigation", aria-label="main navigation", style="background-color:#ff0202;", align="center">
<a href="./index.html" class="navbar-item" style="font-weight: bold; text-decoration: none;background-color:transparent;" align="center">
<img src="https://cdn-icons-png.flaticon.com/512/954/954591.png" alt="Behavior in the Wild" style="width:20px;height:20px;margin-right:5px;">
<b style="color:white;font-weight:bold;">Behavior in the Wild</b>
</a>
</nav>
</header>
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-1 publication-title">Long-Term Ad Memorability</h1>
<h3 class="title is-3 publication-title">Understanding and Generating Memorable Ads</h3>
<div class="is-size-5 publication-authors">
<span class="author-block">
<img src="images/adobe-logo.png" alt="Harini Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://harini-si.github.io/" style="color:#f68946;font-weight:normal;">S I Harini<sup>*</sup></a>,
</span>
<span class="author-block">
<img src="images/adobe-logo.png" alt="Somesh Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://someshsingh22.github.io" style="color:#f68946;font-weight:normal;">Somesh Singh<sup>*</sup></a>,
</span>
<span class="author-block">
<img src="images/adobe-logo.png" alt="Yaman Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://sites.google.com/view/yaman-kumar/" style="color:#f68946;font-weight:normal;">Yaman Kumar<sup>*</sup></a>,
</span>
<span class="author-block">
<img src="images/adobe-logo.png" alt="Aanisha Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://www.linkedin.com/in/aanisha-bhattacharyya/" style="color:#f68946;font-weight:normal;">Aanisha Bhattacharyya</a>,
</span>
<span class="author-block">
<img src="images/BITS_Pilani-Logo.png" alt="Veeky Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://www.bits-pilani.ac.in/goa/veeky-baths/" style="color:#f68946;font-weight:normal;">Veeky Baths</a>,
</span>
<span class="author-block">
<img src="images/ub-logo.png" alt="Changyou Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://cse.buffalo.edu/~changyou/" style="color:#f68946;font-weight:normal;">Changyou Chen</a>,
</span>
<span class="author-block">
<img src="images/iiitd-logo.png" alt="Rajiv Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://www.iiitd.ac.in/rajivratn" style="color:#f68946;font-weight:normal;">Rajiv Ratn Shah</a>,
</span>
<span class="author-block">
<img src="images/adobe-logo.png" alt="Balaji Logo" style="width:20px;height:20px;margin-right:5px;">
<a href="https://scholar.google.com/citations?user=n8iUBg8AAAAJ" style="color:#f68946;font-weight:normal;">Balaji Krishnamurthy</a>,
</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><b style="color:#f68946; font-weight:normal">▶ </b><a href="https://main--dx-portal--adobe.hlx.page/researchers/about" target="_blank">Media and Data Science Research (MDSR) Lab Adobe</a></b></span>
</div>
<div class="is-size-6 publication-authors">
<span class="author-block"><sup>*</sup>Equal Contribution</span>
</div>
<p>Contact <a href="mailto:behavior-in-the-wild@googlegroups.com">behavior-in-the-wild@googlegroups.com</a> for questions and suggestions</p>
<div class="column has-text-centered">
<div class="publication-links">
<span class="link-block">
<a href="https://arxiv.org/abs/2309.00378" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-arxiv"></i>
</span>
<span>arxiv</span>
</a>
</span>
<span class="link-block">
<a href="https://github.com/behavior-in-the-wild/ad-memorability" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code</span>
</a>
</span>
<span class="link-block">
<a href="https://huggingface.co/datasets/behavior-in-the-wild/LAMBDA"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-database"></i>
</span>
<span>LAMBDA Dataset</span>
</a>
</span>
<span class="link-block">
<a href="https://huggingface.co/datasets/behavior-in-the-wild/UltraLAMBDA"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-database"></i>
</span>
<span>UltraLAMBDA Dataset</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<h4 class="subtitle has-text-centered">
🔥<span style="color: #ff3860">[NEW!] </span>We release the first-ever long term ad memorability dataset, LAMDBA, featuring data from 1749 participants and 2205 ads across 276 brands. <br>
🔥<span style="color: #ff3860">[NEW!] </span>Introducing Henry, our novel model for predicting content memorability. Henry achieves state-of-the-art performance across major memorability datasets and demonstrates superior generalization in zero-shot scenarios on unseen datasets.
<br>
🔥<span style="color: #ff3860">[NEW!] </span>We develop SEED (Self rEwarding mEmorability Modeling), a scalable method to generate high-quality memorable ads using automatically annotated data. Ads generated by our method show a 44% higher memorability score than the original ads.<br>
<br>
</h4>
</div>
</div>
</section>
<div style="text-align: center;">
<h5>Take a look at some ad generations by Henry, visualized by <a href="https://www.adobe.com/products/firefly.html" target="_blank">FireFly</a></h5>
<h5>The raw ads generated in text are given in the textblock below each ad.</h5>
<div style="display: flex; justify-content: center;">
<figure style="margin: 0 10px; width: 500px;">
<video height="500" width="500" autoplay="" loop="" muted="" playsinline="" class="w-full object-cover" poster="">
<source type="video/mp4" src="images/maytag-2.mp4">
</video>
<figcaption><i>"Maytag Overnight Wash and Dry" for the brand Costco</i></figcaption>
<div style="text-align: center; margin-top: 10px;">
<br>
<h5>Raw generation</h5>
<textarea style="width: 100%; height: 200px; border: 1px solid #f4f3f1; background-color: #f8f8f8; font-family: Arial, sans-serif; border-radius: 5px; text-align: left;" readonly>
Input Title: "Maytag Overnight Wash and Dry" for the brand Costco
Generated Ad:
The ad is shot in landscape orientation, at a medium pace.
Scene 1:
The scene shows a person standing in front of a washer and dryer.
Emotions: comforting, comfortable
Tags: appliance, laundry, laundry room, man
Voiceover: "Maytag has taken the laundry room by storm with their overnight wash and dry."
Colors: Gray, Off-White, Silver, White, Black, Dark Gray
Scene 2:
The scene shows a person lying on a bed pressing a button on a cell phone.
Emotions: gift, birthday, holiday
Tags: button, catch, smartphone, hand
Voiceover: "Imagine going to bed with dirty clothes and waking up to a fresh, clean load. All without leaving the house.
It's like having your own personal laundry assistant."
Colors: Gray, Silver, Black, Dark Gray, Dark Brown
Scene 3:
The scene shows a man and woman embracing while a washer and dryer are in the background.
Emotions: love, happiness, togetherness
Tags: couple, hug, laundry, laundry room
Voiceover: "Never fight over the laundry room again."
Colors: Gray, Off-White, Silver, White, Black, Dark Gray
Scene 4:
The scene shows a logo of Maytag.
Emotions: trust, confidence
Tags: logo, Maytag
Voiceover: "Get a Maytag and enjoy the luxury of time."
Colors: Gray, Off-White, Silver, White, Black, Dark Gray
</textarea>
</div>
</figure>
<figure style="margin: 0 10px; width: 500px;">
<video height="500" width="500" autoplay="" loop="" muted="" playsinline="" class="w-full object-cover" poster="">
<source type="video/mp4" src="images/brainly-4.mp4">
</video>
<figcaption><i>"Brainly Keep Learning 30sec Final 16x9" for the brand Brainly</i></figcaption>
<div style="text-align: center; margin-top: 10px;">
<br>
<h5>Raw generation</h5>
<textarea style="width: 100%; height: 200px; padding: 10px; border: 1px solid #f9f8f6; background-color: #f8f8f8; font-family: Arial, sans-serif; border-radius: 5px; text-align: left;" readonly>
Input Title: "Brainly Keep Learning 30sec Final 16x9" for the brand Brainly
Generated Ad:
The ad is shot in landscape orientation, at a medium pace.
Scene 1:
The scene shows a girl holding her book.
Emotions: confused, motivational
Tags: woman, book, bag, homework
Voiceover: "What's up? Need help with physics? "
Colors: Yellow, Green, Pink, Silver
Scene 2:
The scene shows a woman sitting at a table and writing.
Emotions: hard-working, positive
Tags: binder, chair, computer, computer desk, table, drawer, girl, homework, laptop, notebook, notepad, pen, pencil, sit, stool, woman, write
Voiceover: "I'm here to help. I'm Brainly."
Colors: Off-White, White, Gray, Silver
Scene 3:
The scene shows a man working on his laptop.
Emotions: togetherness, positive
Tags: binder, chair, computer, computer desk, table, drawer, laptop, notebook, notepad, pen, pencil, sit, stool,man, write
Voiceover: "I'll find the information and show you the steps. "
Colors: Off-White, White, Gray, Silver
Scene 4:
The scene shows a woman teaching on a white board.
Emotions: positive
Tags: whiteboard, girl, homework, laptop, stick, woman, write
Voiceover: "When you have questions, just ask"
Colors: Off-White, White, Blue
Scene 5:
The scene shows a boy sitting on a desk with his hand raised.
Emotions: confidence, happy
Tags: binder, chair, table, drawer, laptop, notebook, notepad, pen, pencil, sit, stool,man, write
Voiceover: "So you can learn from the process and feel good about learning"
Colors: Green, Brown, White, Gray, Silver
Scene 6:
The scene shows the Brainly logo with a child.
Emotions: trust, confidence
Tags: logo, Brainly
Voiceover: "I'm here to help you keep learning. Try it out"
Colors: Off-White, White, Blue
</textarea>
</div>
</figure>
</div>
<div style="clear: both;"></div>
</div>
<section class="section" style="background-color:#efeff081">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
<p>
Marketers spend billions of dollars on advertisements, but to what end? At purchase time, if customers cannot recognize the brand for which they
saw an ad, the money spent on the ad is essentially wasted. Despite its importance in marketing,
until now, there has been no study on the memorability of ads in the ML literature. All previous
memorability studies have been conducted on shortterm recall on specific content types like object
and action videos. On the other hand, the advertising industry only cares about long-term memorability, and ads are almost always highly multimodal. Therefore, we release the first memorability
dataset, <strong>LAMDBA</strong>, consisting of 1749 participants
and 2205 ads covering 276 brands. Running statistical tests over different participant subpopulations
and ad types, we find many interesting insights into
what makes an ad memorable, e.g., fast-moving
ads are more memorable than those with slower
scenes; people who use ad-blockers remember a
lower number of ads than those who don’t. Next,
we present a novel model, <strong>Henry</strong>, to predict the
memorability of a content which achieves state-ofthe-art performance across all prominent literature
memorability datasets. Henry shows strong generalization performance with better results in 0-shot
on unseen datasets. Finally, with the intent of memorable ad generation, we present a scalable method
to build a high-quality memorable ad generation
model by leveraging automatically annotated data.
Our approach, <strong>SEED(Self rEwarding mEmorability Modeling)</strong> , starts with a language model trained
on LAMBDA as seed data and progressively trains
the LLM to generate more memorable ads. We
show that the generated advertisements have <strong>44% higher memorability</strong>
scores than the original ads.
Further, we release a large-scale ad dataset, <strong>UltraLAMBDA</strong>, consisting of <strong>5 million ads</strong> with their
automatically-assigned memorability scores.
</p>
</div>
</div>
</div>
</div>
</section>
<br><br>
<section class="section">
<div class="container is-max-desktop">
<h2 class="title is-3">Interesting Trends in Memorability</h2>
<div class="content">
<ul class="memorable-trends-list">
<li>Food, entertainment, and tech industries have the most memorable ads.</li>
<li>The length of an ad has minimal impact on its memorability.</li>
<li>Ads with rapid scene transitions are more memorable than slower-paced counterparts.</li>
<li>High-level features like object and scene semantics play a crucial role in ad recall. Ads showcasing human images are significantly more memorable than those featuring objects.</li>
<li>Ads evoking negative emotions tend to stay in memory longer than those with positive sentiments.</li>
</ul>
</div>
<figure>
<img src="images/graphs.png" alt="Memorability Trends" style="width: 100%;">
<figcaption><i>Correlations between content factors (a-d), interaction factors (e-g), and customer behavior factors (h-j) with memorability
on LAMBDA samples.</i></figcaption>
</div>
</section>
<section class="section" id="Examples">
<div class="container">
<h2 class="title is-3">Examples from LAMBDA</h2>
The score indicates the memorability of the ad, with 1.0 being the most memorable and 0.0 being the least memorable.<br><br>
<div class="video-panel">
<div class="video-item">
<iframe width="350" height="269" src="https://www.youtube.com/embed/xXojnCfcPTc" frameborder="0" allowfullscreen></iframe>
<div class="video-caption">Score: 0.85</div>
</div>
<div class="video-item">
<iframe width="350" height="269" src="https://www.youtube.com/embed/2ltpJD5WKvM" frameborder="0" allowfullscreen></iframe>
<div class="video-caption">Score: 1.0</div>
</div>
<div class="video-item">
<iframe width="350" height="269" src="https://www.youtube.com/embed/b6Ej0FuVOzw" frameborder="0" allowfullscreen></iframe>
<div class="video-caption">Score: 0.71</div>
</div>
<div class="video-item">
<iframe width="350" height="269" src="https://www.youtube.com/embed/KgxcWoc6F5k" frameborder="0" allowfullscreen></iframe>
<div class="video-caption">Score: 0.57</div>
</div>
<div class="video-item">
<iframe width="350" height="269" src="https://www.youtube.com/embed/oAWrnrjAdTc" frameborder="0" allowfullscreen></iframe>
<div class="video-caption">Score: 0.33</div>
</div>
<div class="video-item">
<iframe width="350" height="269" src="https://www.youtube.com/embed/t8EPhQhNQCU" frameborder="0" allowfullscreen></iframe>
<div class="video-caption">Score: 0.16</div>
</div>
<!-- Add more video items as needed -->
</div>
</section>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>@misc{s2024longtermadmemorabilityunderstanding,
title={Long-Term Ad Memorability: Understanding and Generating Memorable Ads},
author={Harini S I au2 and Somesh Singh and Yaman K Singla and Aanisha Bhattacharyya and Veeky Baths and Changyou Chen and Rajiv Ratn Shah and Balaji Krishnamurthy},
year={2024},
eprint={2309.00378},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2309.00378}}</code></pre>
</div>
</section>
<section class="section" id="TermsOfService">
<div class="container is-max-desktop content">
<h2 class="title">Terms Of Service</h2>
<p>
LAMBDA and UltraLAMBDA are sourced from brand videos available on YouTube, Facebook Ads, and CommonCrawl. The dataset annotations and video links for both LAMBDA and UltraLAMBDA will be released under the MIT License. The videos themselves are subject to the original creators’ licenses. These datasets do not reveal any identities of the annotators or specific individuals. LAMBDA, being curated manually, ensures that no offensive videos are included; however, UltraLAMBDA, sourced from the internet, may contain noisier content. While the videos originate from brands, some brand content may be perceived as offensive by certain individuals.
</p>
</div>
</section>
<section class="section" id="Acknowledgement">
<div class="container is-max-desktop content">
<h2 class="title">Acknowledgement</h2>
<p>
We thank Adobe for their generous sponsorship.
<br>
We thank the LLaMA team for giving us access to their models, and open-source projects, including Vicuna.
</p>
</div>
</section>
<footer class="footer">
<div class="content has-text-centered">
<p>
<strong>Long-Term Ad Memorability</strong> by <a href="https://behavior-in-the-wild.github.io">Behavior in the Wild</a>.
</p>
</div>
</footer>
</body>
</html>