Deep learning based Intelligent Emotion Recognition and Classification System

Khan, Misha Urooj; Abbasi, Muhammad Abdullah; Saeed, Zubair; Asif, Muneeb; Raza, Ali; Urooj, Urooj

doi:10.1109/fit53504.2021.00015

Cited by 7 publications

(5 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This extensive connection mitigates the vanishing gradient issue that deep neural networks often encounter and allows optimal parameter reuse. Additionally, DenseNet architectures can be improved for feature extractors for a variety of image recognition tasks [25][26][27] because they have already been pre-trained on large datasets, i.e., ImageNet. By removing the fully connected layers and substituting identity functions, the DenseNet121 model is used in the given study as a feature extractor which turns it into a feature extraction module.…”

Section: Features Extractionmentioning

confidence: 99%

Cancerous and Non-Cancerous MRI Classification Using Dual DCNN Approach

Saeed,

Bouhali,

et al. 2024

Bioengineering

Self Cite

View full text Add to dashboard Cite

Brain cancer is a life-threatening disease requiring close attention. Early and accurate diagnosis using non-invasive medical imaging is critical for successful treatment and patient survival. However, manual diagnosis by radiologist experts is time-consuming and has limitations in processing large datasets efficiently. Therefore, efficient systems capable of analyzing vast amounts of medical data for early tumor detection are urgently needed. Deep learning (DL) with deep convolutional neural networks (DCNNs) emerges as a promising tool for understanding diseases like brain cancer through medical imaging modalities, especially MRI, which provides detailed soft tissue contrast for visualizing tumors and organs. DL techniques have become more and more popular in current research on brain tumor detection. Unlike traditional machine learning methods requiring manual feature extraction, DL models are adept at handling complex data like MRIs and excel in classification tasks, making them well-suited for medical image analysis applications. This study presents a novel Dual DCNN model that can accurately classify cancerous and non-cancerous MRI samples. Our Dual DCNN model uses two well-performed DL models, i.e., inceptionV3 and denseNet121. Features are extracted from these models by appending a global max pooling layer. The extracted features are then utilized to train the model with the addition of five fully connected layers and finally accurately classify MRI samples as cancerous or non-cancerous. The fully connected layers are retrained to learn the extracted features for better accuracy. The technique achieves 99%, 99%, 98%, and 99% of accuracy, precision, recall, and f1-scores, respectively. Furthermore, this study compares the Dual DCNN’s performance against various well-known DL models, including DenseNet121, InceptionV3, ResNet architectures, EfficientNetB2, SqueezeNet, VGG16, AlexNet, and LeNet-5, with different learning rates. This study indicates that our proposed approach outperforms these established models in terms of performance.

show abstract

Section: Features Extractionmentioning

confidence: 99%

Cancerous and Non-Cancerous MRI Classification Using Dual DCNN Approach

Saeed,

Bouhali,

et al. 2024

Bioengineering

Self Cite

View full text Add to dashboard Cite

show abstract

“…By using the learned parameters from a pre-trained model, the new model can start with a better initialization than random initialization, which can result in faster convergence during training and better performance on the new task as the pre-trained model has already learned rich feature representations from a large dataset such as ImageNet [17]. While facial expression recognition and classification is a challenging task in computer vision due to variations in facial features and lighting conditions [13], transfer learning has demonstrated tremendous potential to overcome the problem in this domain [11,12,18,19]. By leveraging the learned features and parameters of pre-trained models, transfer learning can generate the models with acceptable accuracy via training through limited facial expression images.…”

Section: Cnn Modelsmentioning

confidence: 99%

“…Akhand et al proposed a method based on transfer learning of pre-trained AlexNet architecture on the ImageNet dataset for FER, which reached an accuracy of 70.52% [11]. Similarly, Khan et al applied MobileNetV2-based transfer learning to improve the performance of automatic facial emotion recognition systems, which achieved 98.70% accuracy [12]. While ImageNet is not specifically designed for emotion classification, transfer learning from it to expression recognition can achieve good accuracy with limited data and parameter modifications.…”

Section: Introductionmentioning

confidence: 99%

Interpretability Study of Transfer Learning for Emotion Classification

Shao

2023

ACE

View full text Add to dashboard Cite

Transfer learning is a powerful technique for improving the performance of deep neural networks on a target task with limited labeled data. However, it is not always clear why pre-trained models on unrelated tasks, such as ImageNet, can be effective for emotion recognition tasks. This paper evaluates the performance of transfer learning for facial expression recognition tasks and investigates the influence of network depth on performance using three different typical pre-trained models: VGG16, MobileNet, and ResNet50. The layers of the pre-trained models are frozen, and their output layer is substituted with a new output layer that corresponds to the seven-class emotion classes in the target dataset. The study demonstrates that transfer learning with pre-trained models, even though with limited labelled dataset regarding facial expressions, can be an effective approach for improving the performance of deep neural networks on emotion recognition tasks. The findings provide insights into the interpretability of transfer learning for facial expression recognition tasks and highlight the importance of fine-tuning pre-trained models for optimal performance. Additionally, the study also offers ideas for future model design in facial expression recognition tasks.

show abstract

“…Digital images, videos, and other visual inputs can be used to extract information used by computers and systems in computer vision. These systems respond in response to the input [35,36,37]. Aging affects humans' lives severely.…”

Section: Computer Vision Based Fall Detectionmentioning

confidence: 99%

Fall Detection, Wearable Sensors & Artificial Intelligence: A Short Review

Ishtiaq

Saeed

Khan

et al. 2022

JAREE

View full text Add to dashboard Cite

Falls are a major public health concern among the elderly and the number of gadgets designed to detect them has increased significantly in recent years. This document provides a detailed summary of research done on fall detection systems, with comparisons across different types of studies. Its purpose is to be a resource for doctors and engineers who are planning or conducting field research. Following the examination, datasets, limitations, and future imperatives in fall detection were discussed in detail. The quantity of research using context-aware approaches continues to rise, but there is a new trend toward integrating fall detection into smartphones, as well as the use of artificial intelligence in the detection algorithm. Concerns with real-world performance, usability, and reliability are also highlighted.

show abstract

Deep learning based Intelligent Emotion Recognition and Classification System

Cited by 7 publications

References 19 publications

Cancerous and Non-Cancerous MRI Classification Using Dual DCNN Approach

Cancerous and Non-Cancerous MRI Classification Using Dual DCNN Approach

Interpretability Study of Transfer Learning for Emotion Classification

Fall Detection, Wearable Sensors & Artificial Intelligence: A Short Review

Contact Info

Product

Resources

About