https://github.com/anidipta/task-1

Compare the performance of time-series models with static image-based approaches for emotion recognition.
https://github.com/anidipta/task-1

Last synced: 7 days ago
JSON representation

Compare the performance of time-series models with static image-based approaches for emotion recognition.

• Host: GitHub
• URL: https://github.com/anidipta/task-1
• Owner: Anidipta
• License: apache-2.0
• Created: 2024-04-23T14:55:24.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2024-08-19T16:39:37.000Z (5 months ago)
• Last Synced: 2024-08-19T19:45:27.347Z (5 months ago)
• Language: Jupyter Notebook
• Size: 1.58 MB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        

# 📊 Task-1: Comparing Time-Series Models with Static Image-Based Approaches for Emotion Recognition

### 🎯 Objective:

Compare the performance of time-series models with static image-based approaches to determine which method is more effective for emotion recognition.

### 📂 Project Files:

- **`README.md`** - Updated with project overview and instructions.

- **`Task 1 report.pdf`** - Detailed report of Task 1, summarizing the findings and comparisons.

- **`null-1 (2).ipynb`** - Jupyter notebook with code and analysis for Task 1.

- **`.gitignore`** - Git ignore file to exclude unnecessary files from version control.

- **`LICENSE`** - Project license for usage rights.

### 📈 Performance Metrics:

- **Accuracy Inception:** ![Accuracy Inception](Images/Accuracy Inception.jpg)

- **Accuracy ResNet:** ![Accuracy ResNet](Images/Accuracy Resnet.jpg)

- **Loss Inception:** ![Loss Inception](Images/Loss Inception.jpg)

- **Loss ResNet:** ![Loss ResNet](Images/Loss Resnet.jpg)

- **Confusion Matrix Inception FER:** ![Confusion Matrix Inception FER](Images/confusion matrix inception FER.png)

- **Confusion Matrix ResNet FER:** ![Confusion Matrix ResNet FER](Images/confusion matrix RESNET FER.png)

- **ROC AUC Inception FER:** ![ROC AUC Inception FER](Images/ROC AUC inception FER.png)

- **ROC AUC ResNet FER:** ![ROC AUC ResNet FER](Images/ROC AUC RESNET FER.png)

- **Dataset Distribution:** ![Dataset Distribution](Images/Dataset Distribution.jpg)

### 🎥 Demo Video:

Watch the demo video showcasing the performance comparison of the models:

[View Demo Video](https://github.com/Anidipta/Task-1/assets/140332086/b8623542-d0af-4ff7-9797-ec25d8f738d8)

### 🔍 Project Highlights:

- **Model Comparison:** Evaluates the effectiveness of time-series models vs. static image-based approaches for emotion recognition.

- **Visual Analytics:** Detailed performance metrics and visualizations to compare the models' accuracy, loss, and ROC AUC scores.

- **Insights:** Provides insights into the strengths and weaknesses of each approach in different scenarios.

Feel free to explore the provided files, images, and demo video to understand the comparative performance and the effectiveness of the models used.