OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Emotion Recognition with CNN

A Python and TensorFlow research project that uses a convolutional neural network to classify facial images into seven emotion labels with FER-2013. The repository documents data preparation, model training and an attempted live-use mode, while warning that its code may not run.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Emotion Recognition with CNN — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it as a research or teaching reference when examining a FER-2013 CNN workflow, while accounting for the repository’s warning that the code may not run.

Who it is for

  • Learners exploring CNN-based facial-expression classification.
  • Developers reviewing a FER-2013 data-preparation and training example.

Documented uses

  • Convert the FER-2013 CSV into image and label data for the project.
  • Explore the documented training workflow for a CNN using an AlexNet-style default architecture.
  • Review the project’s proposed alternative architectures in its paper.
Before you choose

Requirements & limits

The project is publicly visible on GitHub under the MIT license. The README says FER-2013 access must be requested or the dataset can be obtained through its linked Kaggle page.

  • The README explicitly warns that the code may not run.
  • The project reports seven emotion labels and uses AlexNet-style architecture by default.
  • The README cautions that dataset quality and image diversity may limit performance outside the dataset.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Prepare FER-2013 data for the project

Set up the documented Python environment and convert the FER-2013 CSV into image and label data.

Before you start

  • Python 3
  • FER-2013 dataset access, obtained through the dataset access request process or the linked Kaggle dataset
  • The repository files, including requirements.txt and csv_to_numpy.py

Steps

  1. Download fer2013.tar.gz and decompress fer2013.csv into the project’s ./data folder.

  2. Create and activate a Python 3 virtual environment with the README’s virtualenv commands.

  3. Install dependencies with pip install -r requirements.txt.

  4. Run python3 csv_to_numpy.py to generate image and label data in the data folder.

Expected outcome

The documented preparation stage places converted image and label data in the project’s data folder.

Guide limitations

  • The README warns that the code may not run.
  • The project cautions that FER-2013’s data quality and image diversity are limited and that performance may be poor in other applications.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
6 November 2026
Original website title
isseu/emotion-recognition-neural-networks · Official homepage title · retained in its source language
How we review sourcesSuggest a correction