×

Search anything:

109 Deep Learning projects [with source code]

Internship at OpenGenus

Get this book -> Problems on Array: For Interviews and Competitive Programming

In this article, we have listed 109 Deep Learning projects that will help you boost your Portfolio. We have provided resources to explore the project ideas further along with source code. You can do this as a part of your College Project (B.Sc, M.Sc and even PhD) or take it up on your own to boost your portfolio and build your practical skills in this domain.

We have categorized all Deep Learning Projects as Beginners, Intermediate and Advanced. Following is the complete list of 109 Deep Learning Projects:

Deep Learning Projects for Beginners

These are Deep Learning projects that you should be comfortable doing as a beginner. This is perfect for 1st and 2nd year Students pursuing B.Sc in Computer Science.

Intermediate Deep Learning projects

These are Deep Learning Projects that you should add to your portfolio if you have a hold on basic Deep Learning concepts and are in later years in B.Sc or pursuing M.Sc in Computer Science. These projects are super beneficial to enhance your portfolio and land a strong career in leading companies like Google, Apple, Facebook and Microsoft.

  • X-Ray image segmentation using XNet
  • Face Aging using Conditional GANs
  • Fault detection system
  • Language Identification Technique
  • Text Generation
  • Social Distancing Checker
  • Early Detection and Diagnosis (EDD) using RefineDet
  • Chest X-Ray Radiograph
  • Colonoscopy polyp detection and classification
  • Diagnosis of skin diseases
  • Multigrade Brain Tumor Classification
  • Heart disease prediction
  • Breast cancer classification
  • Self-Driving Car
  • Traffic Forecasting
  • Traffic flow speed prediction
  • Citywide Crowd Flows Prediction
  • Bike flow prediction
  • Taxi Demand Prediction
  • Road Travel Time Prediction
  • Smart parking
  • Vehicle Safety Improvement
  • Pedestrian Detection for Automated Driving
  • Real-time determination of earthquake
  • Earthquake Magnitude Estimation: Early warning system
  • flood prediction
  • Fingerprint Classification and Identification
  • Personal mobile sensing
  • Fake News Detection
  • Colouring historic black and white photos
  • Image denoising
  • Generate High Resolution images using SRGAN
  • Image caption generator
  • Face detection
  • Human Pose Estimation
  • Auto text completion
  • Automatic Quality Grading of Mangoes
  • Language Translator
  • Video Summarization
  • Video Subtitles generation
  • Next Frame Prediction
  • Video Compression
  • CCTV survelliance
  • Video recommendation
  • Friend recommendation in social media
  • News recommendation
  • Fake News detection
  • Speech recognition with the emotions
  • Fake currency detection
  • Fuel Automation Using Currency Recognition
  • Landmark detection
  • Human Activity Recognition
  • Cat breed recognition
  • Spam Comment Recognition Based on Wide & Deep Learning
  • Electricity-Theft Detection to Secure Smart Grids
  • Peer-to-Peer Lending
  • Multilingual Speech to Text
  • Text translation
  • Multilingual cyber bullying detection
  • Music Genre Classification
  • Music recommendation system
  • Music Popularity Prediction
  • Dual-track Music Generation
  • Drowsiness detection system
  • Image caption generator
  • Image to Image translation
  • Extracting Text from Images
  • Celebrity Look-alike

Advanced Deep Learning Projects

These are Advanced Deep Learning Projects. These may take up significant effort for couple of months to implement and analyze. These will set your apart from the crowd and help you to rise in the competition.

  • Chatbot
  • Virtual Assistant
  • Deep Fake Video Generation
  • Forensic Detection of Fake Image
  • Forensic Verification and Detection of Fake Video
  • Thief Detection
  • Suspicious human activity detection
  • Identifying Missing People
  • Real Time Crime Detection with Video Surveillance
  • ATM Robbery prevention
  • Smart Home Speech Recognition
  • Classification and Diagnosis of Alzheimer's Disease
  • Detection of hate speech
  • Identify Not Suitable for Work (NSFW) Images
  • Body Parts Detection using HRNet and EfficientNet
  • Deep voice: Text to speech
  • Cyberattack detection
  • Secure crowdsensing
  • Differential Privacy
  • Adversial Attacks on a given application
  • Image Recognition using Transfer Learning Approach
  • Analysis of Quantization of Deep Learning models
  • Analysis of Pruned Deep Learning models
  • Optimizing Deep Learning models

Following are further details on some of the above interesting Deep Learning Projects:

1. Early Detection and Diagnosis (EDD)

  • Project: Early Detection and Diagnosis (EDD)
  • Models: RefineDet
  • Datasets: Endoscopy Disease Detection and Segmentation (EDD2020)
  • Application domain: Medical Imaging
  • Level: Beginner
  • Audience Interest level: High
  • Explanation: article on EDD project, paper on CXR-RefineDet
  • Source code: EDD2020

In this project, we have an image of a sample (body part) and we have to develop a Deep Learning model to detect the possible disease. This is an important application as it can help eliminate human error and enable early diagonis of critical ailments.

This can be done as a 2D project which involve images or X-rays or as a 3D project which will involve video like clonoscope.

As a touch of innovation, one can develop a Deep Learning model to predict how the image will look if a particular course of treatment is taken or not taken.

Early Detection and Diagnosis

2. Text Generation

  • Project: Text Generation
  • Models: BERT Large, LSTM, GPT
  • Datasets: Portuguese text generator Dataset
  • Application domain: NLP
  • Level: Moderate
  • Audience Interest level: High
  • Explanation: Using LSTM
  • Source code: BERT text generation

This is one of the most popular applications of Deep Learning specially after the introduction of ChatGPT which reached 1M active users in just 6 days.

3. Object Detection

In this project, one needs to train a Deep Learning model to identify all objects in a given image or video. This is a standard application but can be applied in innovative ways.

As a touch of innovation, one can leverage the technology of Object Detection and try to predict if someone has modified the image. Use case can be a person who takes video of a room before leaving and takes a video again to check if things have been touched.

This is real-time Object Detection:

Object Detection

You can extend this to identify only faces in a crowd:

Face detection

4. Image Recognition

  • Project: Image Recognition
  • Models: ResNet50
  • Datasets: ImageNet
  • Application domain: Image Processing
  • Level: Beginner
  • Audience Interest level: Low
  • Explanation: Image Recognition using Transfer Learning
  • Source code:

This is the most common Deep Learning project but it can be made interesting but incorporating state-of-the-art techniques like Transfer Learning. A touch of innovation will be to extend this to make a Deep Learning model to just distinguish between 2 objects (dog or cat).

Image Recognition

5. Text Summarization

  • Project: Text Summarization
  • Models: BERT, RNN, Transformer
  • Datasets: Any NLP dataset
  • Application domain: Natural Language Processing (NLP) + DL
  • Level: Beginner
  • Audience Interest level: Moderate
  • Explanation: Different techniques, Using Transformers, Using RNN, Using BERT
  • Source code: bert-text-summarizer

In this project, you need to summarize a text into a specified number of words. There are mainly two types: extractive and abstract. In extractive, the main sentences are added to the summary only while in abstractive, the summary is written by the Deep Learning from scratch.

This is an extensively researched topic and is equally interesting. As a touch of innovation, you can couple this with an application like reading terms and conditions.


With this article at OpenGenus, you must have a strong idea of different project ideas you can do for Deep Learning and boost your portfolio.

109 Deep Learning projects [with source code]
Share this