2019 14th IEEE International Conference on Automatic Face &Amp; Gesture Recognition (FG 2019) 2019
DOI: 10.1109/fg.2019.8756614
|View full text |Cite
|
Sign up to set email alerts
|

Deformable Synthesis Model for Emotion Recognition

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
12
0

Year Published

2019
2019
2023
2023

Publication Types

Select...
4
2
1

Relationship

0
7

Authors

Journals

citations
Cited by 15 publications
(12 citation statements)
references
References 25 publications
0
12
0
Order By: Relevance
“…BU-4DFE is widely used on emotion recognition [30] and 3D face alignment [31]. In the test on BU-4DFE, we first used 2,340 reconstructed 3D faces from 36 sequences of facial [29]; column (c) depicts the 3D face reconstructed from our pipeline.…”
Section: Results On 3d Face Reconstructionmentioning
confidence: 99%
“…BU-4DFE is widely used on emotion recognition [30] and 3D face alignment [31]. In the test on BU-4DFE, we first used 2,340 reconstructed 3D faces from 36 sequences of facial [29]; column (c) depicts the 3D face reconstructed from our pipeline.…”
Section: Results On 3d Face Reconstructionmentioning
confidence: 99%
“…The pooling mechanism is requisite for the successful outcomes of CCNs. During this study, it is noted that, if we use convolutional layers with lower strides (1,2) in the place of pooling layers, the accuracy is decreased. However, using convolutions with higher strides (4,5) does not converge the model while the number of parameters are also increased.…”
Section: Effect Of Depth Reduction On the Accuracy Of The Exnetmentioning
confidence: 90%
“…Emotion classification is an essential feature of human intelligence which enables us to communicate appropriately. For communication between humans and machines, the role of emotion classification is even more influential [1]. Expression classification has a wide range of real-world applications such as autism detection, depression detection, retail shopping (customer satisfaction), educational sectors, and pain assessment [2].…”
Section: Introductionmentioning
confidence: 99%
“…Based on a CNN fine-tuned on FER datasets, the IA-gen achieves accuracies of 96.75% on CK+ and 88.92% on Oulu-CASIA. Fabiano et al [ 21 ] first utilized a deep-learning-based 3D morphable model to synthesize 3D images with different expressions, and then employed the synthesize 3D images to train deep neural networks for facial emotion recognition on 3D datasets.…”
Section: Related Workmentioning
confidence: 99%