[Virtual Presenter] We are delighted to welcome you all to our customer training and onboarding presentation. During this session, we will discuss the motivation behind gaze tracking, literature review, and pipeline for predicting engagement patterns. We are thrilled to share our groundbreaking approach to advertising and believe gaze tracking and deep learning hold the key to transforming the advertising landscape. We hope this presentation will be informative and engaging, and we are eager to respond to any questions you may have.. 

Scene 1

Close-up of a person's eye

Description automatically generated

FocusFrame: Enhancing Advertising Optimization 
by Predicting Engagement Through Gaze Tracking 
Technology and Deep Learning

Presented by:   Laiba Meer      400903 Supervisor:        Dr. Wajahat Hussain

[Audio] We are here to discuss the power of gaze tracking technology and deep learning in predicting engagement and optimizing advertising. Gaze tracking technology enables us to precisely measure visual attention, allowing us to understand viewer focus and develop effective ad strategies. With deep learning, we can analyze this data to determine what engages viewers and adjust our ads accordingly. This combination of technologies allows us to create ads that are not only effective but also tailored to the specific needs and interests of our audience. Whether you are a marketer or an advertiser, gaze tracking technology and deep learning can help you take your advertising to the next level.. 

Scene 2

A collage of images of a person and a bottle of suntan lotion

Description automatically generated

A person with a colorful circle

Description automatically generated with medium confidence

A collage of food

Description automatically generated

Understanding viewer focus is key for effective ad design Eye-tracking precisely measures visual attention for better strategies

A yellow bag with a picture of potato chips Description automatically generated. 

Scene 3

A yellow bag with a picture of potato chips

Description automatically generated

A screenshot of a product

Description automatically generated

[Audio] We are here to discuss the use of gaze tracking and deep learning in predicting engagement and optimizing advertising. We believe that combining gaze tracking and deep learning can accurately predict where a user will see on graphical interfaces and optimize advertising to better engage and convert users. But why gaze tracking? Gaze tracking is a technology that enables us to track where a user's eyes are looking at on graphical interfaces. This information can be used to understand how a user interacts with the interface and make changes to improve engagement and conversion. We use gaze tracking to understand how a user interacts with the interface and make changes to improve engagement and conversion. By combining gaze tracking and deep learning, we can accurately predict where a user will see on graphical interfaces and optimize advertising to better engage and convert users. Thank you for your time and attention.. 

Scene 4

To predict where a user will see on graphical interfaces

 Integration of Gaze tracking and Deep learning

 Why Gaze Tracking?

[Audio] In this presentation, we will discuss our use of gaze tracking technology and deep learning to predict advertising performance and improve outcomes. We will provide examples of successful applications of gaze tracking in the advertising industry.. 

Scene 5

Motivation: (1) Success Stories with Gaze Tracking

A screenshot of a website

Description automatically generated

Alka Forsikring increased email CTR by 63% using Neurons’ AI attention prediction tool

[Audio] We use gaze tracking technology and deep learning to predict customer engagement and optimize advertising. Understanding customer engagement is crucial for the success of advertising campaigns. By analyzing customer behavior, we can gather insights on what is working and what is not. This information can then be used to optimize advertising, ensuring that it reaches the right audience and achieves the desired results. Our gaze tracking technology allows us to gather detailed data on customer engagement with advertising, which can help to improve engagement and increase sales. By tailoring advertising to meet the needs of customers, we can increase the effectiveness of advertising campaigns and build stronger relationships with customers. Our gaze tracking technology and deep learning capabilities provide a powerful tool for predicting customer engagement and optimizing advertising. By working with us, you can ensure that your advertising is reaching the right audience and achieving the desired results.. 

Scene 6

A screenshot of a phone

Description automatically generated

A computer screen with a receipt coming out of it

Description automatically generated

NEOHUB increased fastPay’s click-through rates by 65% using predictive AI

[Audio] Good afternoon, everyone. It is our pleasure to present to you our research on utilizing gaze tracking technology and deep learning to enhance user engagement and optimize advertising. Today, we will focus on the study, UEyes: Understanding visual saliency across user interface types. In this study, we explore the potential of eye-tracking data to predict and optimize user attention patterns across various types of user interfaces. Our research suggests that gaze tracking technology, when combined with deep learning algorithms, can offer valuable insights into how users interact with different types of UIs. These insights can be used to improve the design and functionality of UIs, resulting in higher user engagement and satisfaction. Thank you for your time, and we look forward to discussing our findings in greater detail.. 

Scene 7

Jiang, Y., Leiva, L. A., Rezazadegan Tavakoli, H., RB Houssel, P., Kylmälä, J., & Oulasvirta, A. (2023, April). UEyes: Understanding visual saliency across user interface types. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (pp. 1-21). https://doi.org/10.1145/3544548.3581096

How eye-tracking data can be used to predict and optimize user attention patterns across different types of user interfaces (UIs)

[image] GBVS
SAM Pretrained
SAM on UEyes
SAM.. on UEyes
p retrained
UNISI on UEyes
UMSI.• on UEyes
AUC.Judd T
0756 ± 0.104
0.822 0.074
0885 ± 0.057
0868 ± 0.060
0.778 ± 0.090
0.878 ± 0.066
0.905 ± 0.044
NSS T
0.256 ± 0197
0.377 0.170
0.434 ± 0.185
0.414 ± 0.179
0.346 0.178
0.424 ± 0.187
0.401 0.173
3.214 ± 0.668
3.143 ± 0.768
3.337 ± 0.774
3.165 ± 0.774
3.177 ± 0.796
3.376 ± 0.807
3.320 ± 0.744
SIM t
0.513 ± 0.097
(1562 0.081
0663 ± 0.081
0.666 ± 0.080
0.521 0.078
0.639 ± 0.085
0.733 ± 0.069
CCT
0.314 ± 0.193
0.522 ± (1146
0.720 ± 0.127
0.717 ± 0.127
0.431 0.155
0.699 ± 0.156
0.833 ± 0.078
3.916 ± 2.630
2.721 1.457
2.016 ± 1263
1.604 ± 1.185
3.757 1.769
2.676 ± 1.408
1.166 * 0.772
Table I: Saliency maps' quantitative evaluation, With mean ± SD reported for each metric. Arrows indicate the direction Of the
importance; e.g., T means "higher is better". The best result in each column is presented in bold. UMSI+* outperforms the other
models for most evaluation metrics.

[Audio] We are delighted to welcome you to our customer training and onboarding session on our gaze tracking technology and deep learning for advertising. During this session, we will discuss the latest research paper, 'Advertising image saliency prediction method based on score level fusion'. This paper showcases our innovative saliency prediction algorithm, which improves accuracy and inference speed for advertising images by combining text and natural image features. Our gaze tracking technology and deep learning can be used to predict engagement and optimize advertising by identifying which parts of an image are most important to a viewer. This information can be used to create targeted advertising campaigns that resonate better with your audience. By leveraging our gaze tracking technology and deep learning, you can take your advertising to the next level. Our novel saliency prediction algorithm can enhance the accuracy and speed of your advertising campaigns, ensuring that your advertising is engaging and effective. Thank you for joining us in this session, and we look forward to working with you in the future.. 

Scene 8

Kou, Q., Liu, R., Lv, C., Jiang, H., & Cheng, D. (2023). Advertising image saliency prediction method based on score level fusion. IEEE Access, 11, 8455-8466. https://doi.org/10.1109/ACCESS.2023.3236807

How a novel saliency prediction algorithm enhances accuracy and inference speed for advertising images by integrating text and natural image features.

[image] Cornparisons of results in ablation experiments The best and second best results are shown in black bold and blue bold.
TABLE 3.
Pre-trained AMC-SNet [171
Fine-tuned AMC-SNet without central prior
Fine-tuned AMC-SNet with central prior
0.856
0.871
0.717
m726
0.725
2.763
2.663
2.664
0.755
0.771
0.773
1.35
1.29
1.29
0.668
0.690
0.715

[Audio] We will discuss our pipeline approach to predicting engagement and optimizing advertising. Our pipeline consists of several steps that work together to determine the engagement potential of ads. First, we use gaze tracking technology to track where the audience is looking. This helps us understand which parts of the ad are most important to them and which parts they may be ignoring. Second, we apply a deep learning model to analyze the eye fixations recorded during the gaze tracking process. This model helps us identify patterns and trends in the data that can be used to predict which ads are most likely to engage the audience. Third, we apply Gaussian kernel and averaging techniques to the fixation map generated by the gaze tracking process. This helps us refine our predictions and improve the accuracy of our model. Finally, we use a convolutional neural network (C-N-N--) model to make our predictions about which ads are most likely to engage the audience. This model is trained on a large dataset of eye fixation data and can make highly accurate predictions about which ads are most likely to be successful. We hope this has helped you understand our pipeline and how it can be used to predict engagement and optimize advertising. If you have any questions, please feel free to ask. Thank you for your attention.. 

Scene 9

[image] A diagram of a person sitting at a desk Description automatically generated

[image] A diagram of a model of salience Description automatically generated

A black background with a black square

Description automatically generated with medium confidence

[image] A person standing next to a car Description automatically generated

[image] A person walking in a room Description automatically generated

[image] A white spot on a black background Description automatically generated

Results:. [image] A collage of images of clouds Description automatically generated. 

Scene 10

[image] A collage of images of clouds Description automatically generated

[image] A black and white image of a couple of objects Description automatically generated

Cont.. [image] A collage of images of a person s body Description automatically generated. 

Scene 11

[image] A collage of images of a person s body Description automatically generated

[image] A black and white image of different shapes Description automatically generated

[Audio] We aim to enhance the process of predicting engagement and optimizing advertising through the use of gaze tracking technology and deep learning. To achieve this, we propose using lecture slides for attention prediction. By incorporating demographic data into our dataset, we can improve our ability to predict engagement and optimize our advertising efforts.. 

Scene 12

Using lecture slides for attention prediction Expanding the dataset with demographic data.

[Audio] We are pleased to welcome you to our presentation. Today, our focus will be on our use of gaze tracking technology and deep learning to predict engagement and optimize advertising. Our proposed timeline for this project is as follows:. 

Scene 13

Activities Sep-24 Oct-24 Nov-24 Dec-24 Jan-25 Feb-25 Literature Review





 Data Collection





 Model Development





 Inhouse Defense





 Final Defense

[Audio] We have been researching and developing gaze tracking technology and deep learning for years, and we are proud to see the results. Our team has found that gaze tracking technology and deep learning can accurately predict which parts of an ad are most likely to capture the user's attention, and which parts are most likely to lead to user engagement. This is a significant advantage for advertisers, as it means they can create ads that are more likely to be seen and remembered by their target audience. We also found that gaze tracking technology and deep learning can be used to optimize advertising in real time. For example, if we see that a certain part of an ad is not performing well, we can adjust the ad to focus more on that part. This is a significant advantage for businesses, as it means they can get more out of their advertising budget. Overall, we are very excited about the potential of gaze tracking technology and deep learning for advertising. We believe that this technology has the power to revolutionize the way advertisers reach their target audience and drive more engagement.. 

Scene 14

Jiang, Y., Leiva, L. A., Rezazadegan Tavakoli, H., RB Houssel, P., Kylmälä, J., & Oulasvirta, A. (2023, April). UEyes: Understanding visual saliency across user interface types. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (pp. 1-21). Fosco, C., Casser, V., Bedi, A. K., O'Donovan, P., Hertzmann, A., & Bylinskii, Z. (2020, October). Predicting visual importance across graphic design types. In Proceedings of the 33rd Annual ACM Symposium on User Interface Software and Technology (pp. 249-260). neurons. (n.d.). Retrieved from www.neurons.com: https://www.neuronsinc.com/cases/how-alka-increased-ctr-by-63-with-ai-powered-email-optimization Kou, Q., Liu, R., Lv, C., Jiang, H., & Cheng, D. (2023). Advertising image saliency prediction method based on score level fusion. IEEE Access, 11, 8455-8466.

Infinite question marks in 3D rendering. Any Questions. 

Scene 15

We are here to discuss the power of gaze tracking technology and deep learning in predicting engagement and optimizing advertising. Gaze tracking technology enables us to precisely measure visual attention, allowing us to understand viewer focus and develop effective ad strategies. With deep learning, we can analyze this data to determine what engages viewers and adjust our ads accordingly. This combination of technologies allows us to create ads that are not only effective but also tailored to the specific needs and interests of our audience. Whether you are a marketer or an advertiser, gaze tracking technology and deep learning can help you take your advertising to the next level.

avatar

We are here to discuss the use of gaze tracking and deep learning in predicting engagement and optimizing advertising. We believe that combining gaze tracking and deep learning can accurately predict where a user will see on graphical interfaces and optimize advertising to better engage and convert users. But why gaze tracking? Gaze tracking is a technology that enables us to track where a user's eyes are looking at on graphical interfaces. This information can be used to understand how a user interacts with the interface and make changes to improve engagement and conversion. We use gaze tracking to understand how a user interacts with the interface and make changes to improve engagement and conversion. By combining gaze tracking and deep learning, we can accurately predict where a user will see on graphical interfaces and optimize advertising to better engage and convert users. Thank you for your time and attention.

In this presentation, we will discuss our use of gaze tracking technology and deep learning to predict advertising performance and improve outcomes. We will provide examples of successful applications of gaze tracking in the advertising industry.

We use gaze tracking technology and deep learning to predict customer engagement and optimize advertising. Understanding customer engagement is crucial for the success of advertising campaigns. By analyzing customer behavior, we can gather insights on what is working and what is not. This information can then be used to optimize advertising, ensuring that it reaches the right audience and achieves the desired results.  Our gaze tracking technology allows us to gather detailed data on customer engagement with advertising, which can help to improve engagement and increase sales. By tailoring advertising to meet the needs of customers, we can increase the effectiveness of advertising campaigns and build stronger relationships with customers.  Our gaze tracking technology and deep learning capabilities provide a powerful tool for predicting customer engagement and optimizing advertising. By working with us, you can ensure that your advertising is reaching the right audience and achieving the desired results.

Good afternoon, everyone. It is our pleasure to present to you our research on utilizing gaze tracking technology and deep learning to enhance user engagement and optimize advertising. Today, we will focus on the study, UEyes: Understanding visual saliency across user interface types. In this study, we explore the potential of eye-tracking data to predict and optimize user attention patterns across various types of user interfaces. Our research suggests that gaze tracking technology, when combined with deep learning algorithms, can offer valuable insights into how users interact with different types of UIs. These insights can be used to improve the design and functionality of UIs, resulting in higher user engagement and satisfaction. Thank you for your time, and we look forward to discussing our findings in greater detail.

 Jiang, Y., Leiva, L. A., Rezazadegan Tavakoli, H., RB Houssel, P., Kylmälä, J., & Oulasvirta, A. (2023, April). UEyes: Understanding visual saliency across user interface types. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (pp. 1-21). https://doi.org/10.1145/3544548.3581096

slide7-hotlink6

We are delighted to welcome you to our customer training and onboarding session on our gaze tracking technology and deep learning for advertising. During this session, we will discuss the latest research paper, 'Advertising image saliency prediction method based on score level fusion'. This paper showcases our innovative saliency prediction algorithm, which improves accuracy and inference speed for advertising images by combining text and natural image features. Our gaze tracking technology and deep learning can be used to predict engagement and optimize advertising by identifying which parts of an image are most important to a viewer. This information can be used to create targeted advertising campaigns that resonate better with your audience. By leveraging our gaze tracking technology and deep learning, you can take your advertising to the next level. Our novel saliency prediction algorithm can enhance the accuracy and speed of your advertising campaigns, ensuring that your advertising is engaging and effective. Thank you for joining us in this session, and we look forward to working with you in the future.

We will discuss our pipeline approach to predicting engagement and optimizing advertising. Our pipeline consists of several steps that work together to determine the engagement potential of ads. First, we use gaze tracking technology to track where the audience is looking. This helps us understand which parts of the ad are most important to them and which parts they may be ignoring. Second, we apply a deep learning model to analyze the eye fixations recorded during the gaze tracking process. This model helps us identify patterns and trends in the data that can be used to predict which ads are most likely to engage the audience. Third, we apply Gaussian kernel and averaging techniques to the fixation map generated by the gaze tracking process. This helps us refine our predictions and improve the accuracy of our model. Finally, we use a convolutional neural network (C-N-N--) model to make our predictions about which ads are most likely to engage the audience. This model is trained on a large dataset of eye fixation data and can make highly accurate predictions about which ads are most likely to be successful. We hope this has helped you understand our pipeline and how it can be used to predict engagement and optimize advertising. If you have any questions, please feel free to ask. Thank you for your attention.

We aim to enhance the process of predicting engagement and optimizing advertising through the use of gaze tracking technology and deep learning. To achieve this, we propose using lecture slides for attention prediction. By incorporating demographic data into our dataset, we can improve our ability to predict engagement and optimize our advertising efforts.

We are pleased to welcome you to our presentation. Today, our focus will be on our use of gaze tracking technology and deep learning to predict engagement and optimize advertising. Our proposed timeline for this project is as follows:

We have been researching and developing gaze tracking technology and deep learning for years, and we are proud to see the results. Our team has found that gaze tracking technology and deep learning can accurately predict which parts of an ad are most likely to capture the user's attention, and which parts are most likely to lead to user engagement. This is a significant advantage for advertisers, as it means they can create ads that are more likely to be seen and remembered by their target audience. We also found that gaze tracking technology and deep learning can be used to optimize advertising in real time. For example, if we see that a certain part of an ad is not performing well, we can adjust the ad to focus more on that part. This is a significant advantage for businesses, as it means they can get more out of their advertising budget. Overall, we are very excited about the potential of gaze tracking technology and deep learning for advertising. We believe that this technology has the power to revolutionize the way advertisers reach their target audience and drive more engagement.

 Jiang, Y., Leiva, L. A., Rezazadegan Tavakoli, H., RB Houssel, P., Kylmälä, J., & Oulasvirta, A. (2023, April). UEyes: Understanding visual saliency across user interface types. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (pp. 1-21). Fosco, C., Casser, V., Bedi, A. K., O'Donovan, P., Hertzmann, A., & Bylinskii, Z. (2020, October). Predicting visual importance across graphic design types. In Proceedings of the 33rd Annual ACM Symposium on User Interface Software and Technology (pp. 249-260). neurons. (n.d.). Retrieved from www.neurons.com: https://www.neuronsinc.com/cases/how-alka-increased-ctr-by-63-with-ai-powered-email-optimization Kou, Q., Liu, R., Lv, C., Jiang, H., & Cheng, D. (2023). Advertising image saliency prediction method based on score level fusion. IEEE Access, 11, 8455-8466.



FocusFrame: Enhancing Advertising Optimization by Predicting Engagement Through Gaze Tracking Technology and Deep Learning