top of page

If you are curious about artificial intelligence and machine learning, you've come to the right place. We are the first AI blog by-and-for Latin Americans.

Welcome!


Source: 0fjd125gk87 on Pixabay


As the field of artificial intelligence continues to advance, new ethical questions are arising. In light of multiple lawsuits involving Midjourney, an AI art generator, and DeviantArt, a popular art site, the debate is whether AI-generated art should be subject to copyright claims from the artists whose art was used for training the data.


The Case for AI Generated Art

On the one hand, some argue that AI-generated art is a form of original expression and should be protected as such. They argue that the AI system is essentially an artist in and of itself, using the data it was trained on as its inspiration. Additionally, AI-generated art often goes through a significant amount of processing and manipulation before it is finalized, making it distinct from the original artwork.


The Case for Artists

On the other hand, others argue that the artists whose work was used to train the AI should be credited and compensated for the use of their artwork. They argue that without the use of this original artwork, the AI-generated art would not exist. Additionally, they argue that it is unfair for the creators of the original artwork to have their work used without permission or compensation.


Potential Solutions

One solution could be to treat AI-generated art similarly to how we treat cover songs or samples in music. In these cases, the original artist is credited and compensated, but the new artist is also able to create something new and original based on the original work. In the same way, the AI system could be credited as the creator of the new artwork, but the original artists whose work was used for training the data would also be compensated and credited.


Another solution could be to have the AI-generated art to be considered as a transformative use of the original art, this way the artist would be credited but the AI generated work would be protected as a new work.


The question of whether AI-generated art should be subject to copyright claims from the artists whose art was used for training the data is a complex one with valid arguments on both sides. It is important to consider the rights and interests of both the AI creators and the original artists, while also finding a solution that promotes creativity and innovation in the field of AI.


What do you think? Do you believe that AI generated art should be considered fair use, or should artists get the credit for the images generated based on their work?



Adept AI Introduces Action Transformer (ACT-1)


In September 2022, Adept AI, a company that specializes in developing advanced AI models, introduced its latest model, Action Transformer (ACT-1). This model is a large-scale Transformer that is trained to use digital tools, including web browsers. The company believes that this model is a step towards achieving artificial general intelligence (AGI) and that the next era of computing will be defined by natural language interfaces that allow users to tell their computers what they want directly.


Why ACT-1 is Important


ACT-1 represents a significant step towards achieving true general intelligence. It is a foundation model for actions, trained to use a wide range of software tools, APIs, and web applications. Adept AI believes that by creating a model that can understand and execute a wide range of tasks and commands, they will be able to achieve the ambitious goal of creating a system that can do anything a human can do in front of a computer. ACT-1 is the first step in this direction, and the company plans to continue developing and improving the model to bring us closer to this goal.


ACT-1 Capabilities


ACT-1 is currently connected to a Chrome extension which allows it to observe what's happening in the browser and take certain actions, like clicking, typing, and scrolling, etc. The company has released videos showcasing some of the capabilities of ACT-1, including high-level user requests, working in-depth in tools like spreadsheets, completing tasks that require composing multiple tools together, and looking up information online. The company believes that the model's capabilities will improve over time and that future systems will have latency that's largely imperceptible to humans.




Home buyers using ACT-1 to look up homes within their budget. Source: Adept AI




Employees using ACT-1 to log a description for a recent meeting on the company dashboard. Source: Adept AI




Someone using ACT-1 on Excel to add columns to their spreadsheet and automatically perform calculations. Source: Adept AI


Looking Ahead


Adept AI believes that natural language interfaces, powered by action transformers like ACT-1, will dramatically expand what people can do in front of a computer, phone, or internet-connected device. In the future, the company expects most interaction with computers to be done using natural language, not graphical user interfaces. Beginners will become power users, no training required, and software will become even more powerful as advanced AI models like ACT-1 help users implement their ideas in language. Sign-ups are now open for the alpha release of the company's first product built around ACT-1.


Additionally, Adept AI sees ACT-1 as being particularly beneficial for businesses and organizations, where it can automate tedious and repetitive tasks, increasing productivity and efficiency. It can also be used for customer service and support, allowing for faster and more accurate responses to customer inquiries and concerns.


Overall, the introduction of ACT-1 represents a significant step forward in the field of AI and has the potential to greatly impact the way we interact with technology in our everyday lives. Adept AI plans to continue to develop and improve ACT-1, with the goal of creating a truly general AI model that can understand and execute a wide range of tasks and commands.


Learn more about ACT-1 and other Adept AI technologies at their website: https://www.adept.ai/



Whether you're an absolute beginner or advanced AI veteran, we have the project for you! Strengthen your knowledge of machine learning with our project ideas. For each project, we will include the difficulty, estimated time to complete so you can pace yourself, and an online example / tutorial for guidance.


Projects

  1. Project Name: Image Classification Description: Train a deep learning model to classify images of different objects. The model will be trained on a dataset of labeled images and will be able to predict the class of new images. Estimated time to complete: 2-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://www.tensorflow.org/tutorials/images/classification Difficulty level: Intermediate

  2. Project Name: Sentiment Analysis Description: Develop a model that can predict the sentiment (positive or negative) of a given piece of text. The model will be trained on a dataset of labeled text and will be able to predict the sentiment of new text. Estimated time to complete: 1-2 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://www.tensorflow.org/tutorials/text/word_embeddings Difficulty level: Intermediate

  3. Project Name: Voice-controlled Home Automation Description: Create an AI-powered voice assistant that can control various home automation devices using voice commands. The system will be able to understand and respond to various commands. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, Raspberry Pi Online example: https://www.tensorflow.org/lite/models/speech_commands Difficulty level: Advanced

  4. Project Name: Chatbot Description: Build a chatbot that can answer questions and carry out simple tasks. The chatbot will be trained on a dataset of question-answer pairs and will be able to understand and respond to natural language inputs. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://www.tensorflow.org/tutorials/text/transformer Difficulty level: Intermediate

  5. Project Name: Self-driving car simulation Description: Create a simulation of a self-driving car using reinforcement learning. The car will learn to navigate a virtual environment and avoid obstacles. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, Unity Online example: https://github.com/Unity-Technologies/ml-agents Difficulty level: Advanced

  6. Project Name: Face recognition Description: Develop a model that can recognize faces in images and videos. The model will be trained on a dataset of labeled images and will be able to identify new faces. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/ageitgey/face_recognition Difficulty level: Intermediate

  7. Project Name: Language Translation Description: Develop a model that can translate text from one language to another. The model will be trained on a dataset of bilingual text and will be able to translate new text. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://github.com/tensorflow/tensor2tensor Difficulty level: Advanced

  8. Project Name: Object Detection Description: Train a deep learning model to detect objects in images and videos. The model will be trained on a dataset of labeled images and will be able to identify new objects. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/tensorflow/models/tree/master/research/object_detection Difficulty level: Intermediate

  9. Project Name: GAN for Image Generation Description: Implement a Generative Adversarial Network (GAN) to generate new images. The model will be trained on a dataset of images and will be able to generate new images based on the data. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras Online example: https://github.com/tensorflow/tensorflow/tree/master/tensorflow/examples/generative_examples Difficulty level: Advanced

  10. Project Name: Emotion Recognition Description: Develop a model that can recognize emotions in images and videos. The model will be trained on a dataset of labeled images and will be able to identify emotions in new images. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/oarriaga/face_classification Difficulty level: Intermediate

  11. Project Name: Text Summarization Description: Develop a model that can summarize a piece of text. The model will be trained on a dataset of labeled text and will be able to generate summaries of new text. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://github.com/tensorflow/models/tree/master/textsum Difficulty level: Intermediate

  12. Project Name: Music Generation Description: Train a deep learning model to generate music. The model will be trained on a dataset of music and will be able to generate new music based on the data. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, Music21 Online example: https://github.com/Skuldur/Classical-Piano-Composer Difficulty level: Advanced

  13. Project Name: Voice Recognition Description: Develop a model that can recognize speech and transcribe it to text. The model will be trained on a dataset of speech and will be able to transcribe new speech. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, SpeechRecognition Online example: https://github.com/Uberi/speech_recognition Difficulty level: Intermediate

  14. Project Name: Time Series Forecasting Description: Develop a model that can forecast future values of a time series. The model will be trained on a dataset of time series data and will be able to forecast new values. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/tensorflow/models/tree/master/research/timeseries Difficulty level: Intermediate

  15. Project Name: Speech Synthesis Description: Develop a model that can synthesize speech from text. The model will be trained on a dataset of speech and text and will be able to synthesize new speech. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, gTTS Online example: https://github.com/pndurette/gTTS Difficulty level: Intermediate

  16. Project Name: Named Entity Recognition Description: Develop a model that can recognize and classify named entities in text. The model will be trained on a dataset of labeled text and will be able to recognize new named entities. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://github.com/tensorflow/models/tree/master/research/language Difficulty level: Intermediate

  17. Project Name: Image Segmentation Description: Train a deep learning model to segment images into different regions. The model will be trained on a dataset of labeled images and will be able to segment new images. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/tensorflow/models/tree/master/research/deeplab Difficulty level: Intermediate

  18. Project Name: Handwriting Recognition Description: Develop a model that can recognize handwriting in images. The model will be trained on a dataset of labeled images and will be able to recognize new handwriting. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/githubharald/SimpleHTR Difficulty level: Intermediate

  19. Project Name: Image Captioning Description: Develop a model that can generate captions for images. The model will be trained on a dataset of labeled images and captions and will be able to generate captions for new images. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/tensorflow/models/tree/master/research/im2txt Difficulty level: Advanced

  20. Project Name: Natural Language Processing (NLP) Description: Develop a model that can perform natural language processing tasks such as text classification and sentiment analysis. The model will be trained on a dataset of labeled text and will be able to perform NLP tasks on new text. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://github.com/nltk/nltk Difficulty level: Beginner

  21. Project Name: Game Playing AI Description: Develop an AI agent that can play a game. The agent will be trained to make decisions and take actions in the game. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenAI Gym Online example: https://gym.openai.com/ Difficulty level: Intermediate

  22. Project Name: Recommender System for books Description: Develop a model that can recommend books to users based on their reading preferences. The model will be trained on a dataset of user preferences and book information and will be able to suggest new books to users. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/zygmuntz/goodbooks-10k Difficulty level: Beginner

  23. Project Name: Face Swapping Description: Develop a model that can swap faces in images and videos. The model will be trained on a dataset of labeled images and will be able to swap faces in new images and videos. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/deepfakes/faceswap Difficulty level: Advanced

  24. Project Name: Text-to-Speech Description: Develop a model that can convert text to speech. The model will be trained on a dataset of speech and text and will be able to generate new speech based on new text inputs. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, gTTS Online example: https://github.com/pndurette/gTTS Difficulty level: Intermediate

  25. Project Name: Predictive Maintenance Description: Develop a model that can predict when a machine is likely to fail. The model will be trained on a dataset of machine sensor data and will be able to predict when a machine is likely to fail. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/Azure/MachineLearningNotebooks/tree/master/how-to-use-azureml/machine-learning-with-python/predictive-maintenance-modeling Difficulty level: Intermediate

  26. Project Name: Image Inpainting Description: Develop a model that can fill in missing parts of an image. The model will be trained on a dataset of images and will be able to fill in missing parts of new images. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/NVIDIA/partialconv-pytorch Difficulty level: Advanced

  27. Project Name: Object Tracking Description: Develop a model that can track objects in images and videos. The model will be trained on a dataset of labeled images and will be able to track new objects in new images and videos. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/jwyang/faster-rcnn.pytorch Difficulty level: Intermediate

  28. Project Name: Image Denoising Description: Develop a model that can remove noise from images. The model will be trained on a dataset of noisy images and will be able to remove noise from new images. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/SaoYan/DnCNN-PyTorch Difficulty level: Intermediate

  29. Project Name: Stock Price Prediction Description: Develop a model that can predict future stock prices. The model will be trained on a dataset of historical stock data and will be able to predict future stock prices. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/llSourcell/How-to-Predict-Stock-Prices-Easily-Demo Difficulty level: Beginner

  30. Project Name: Image Super-Resolution Description: Develop a model that can increase the resolution of images. The model will be trained on a dataset of images and will be able to increase the resolution of new images. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/twtygqyy/pytorch-SRResNet Difficulty level: Advanced

  31. Project Name: Human Activity Recognition Description: Develop a model that can recognize human activities in images and videos. The model will be trained on a dataset of labeled images and videos and will be able to recognize new activities. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/guillaume-chevalier/LSTM-Human-Activity-Recognition Difficulty level: Intermediate

  32. Project Name: Speech Emotion Recognition Description: Develop a model that can recognize emotions in speech. The model will be trained on a dataset of labeled speech and will be able to recognize emotions in new speech. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, SpeechRecognition Online example: https://github.com/happynoom/Speech-Emotion-Recognition Difficulty level: Intermediate

  33. Project Name: Image Compression Description: Develop a model that can compress images. The model will be trained on a dataset of images and will be able to compress new images. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/tensorflow/models/tree/master/research/compression Difficulty level: Intermediate

  34. Project Name: Anomaly Detection Description: Develop a model that can detect anomalies in data. The model will be trained on a dataset of normal data and will be able to detect anomalies in new data. Estimated time to complete: 2-3 weeks Techstack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/yzhao062/anomaly-detection Difficulty level: Beginner

  35. Project Name: Sentiment Analysis in Social Media Description: Develop a model that can analyze the sentiment of tweets on a specific topic. The model will be trained on a dataset of labeled tweets and will be able to analyze the sentiment of new tweets. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, NLTK, Tweepy Online example: https://github.com/vprusso/youtube_tutorials/tree/master/twitter_python Difficulty level: Intermediate

  36. Project Name: Music Generation with RNN Description: Develop a model that can generate new music using Recurrent Neural Networks (RNNs). The model will be trained on a dataset of music and will be able to generate new music based on the data. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, Music21 Online example: https://github.com/Skuldur/Classical-Piano-Composer Difficulty level: Advanced

  37. Project Name: Fraud Detection in Finance Description: Develop a model that can detect fraudulent transactions in financial data. The model will be trained on a dataset of labeled financial data and will be able to detect fraud in new transactions. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/nsethi31/Kaggle-Fraud-Detection Difficulty level: Intermediate

  38. Project Name: Predictive Maintenance in Healthcare Description: Develop a model that can predict when medical equipment is likely to fail. The model will be trained on a dataset of sensor data from medical equipment and will be able to predict when equipment needs maintenance. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/Azure/MachineLearningNotebooks/tree/master/how-to-use-azureml/machine-learning-with-python/predictive-maintenance-modeling Difficulty level: Intermediate

  39. Project Name: Product Recommendation in Retail Description: Develop a model that can recommend products to customers based on their purchase history. The model will be trained on a dataset of customer purchase data and will be able to suggest new products to customers. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/zygmuntz/goodbooks-10k Difficulty level: Beginner

  40. Project Name: Travel Itinerary Suggestion Description: Develop a model that can suggest travel itineraries based on a user's preferences. The model will be trained on a dataset of travel information and user preferences and will be able to suggest new itineraries to users. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/llSourcell/Travel_Itinerary_Generation Difficulty level: Intermediate

  41. Project Name: Customer Segmentation in Business Description: Develop a model that can segment customers into different groups based on their characteristics and behavior. The model will be trained on a dataset of customer data and will be able to segment new customers. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/nsethi31/Kaggle-Customer-Segmentation Difficulty level: Intermediate

  42. Project Name: Predictive Pricing in Retail Description: Develop a model that can predict the optimal price for a product. The model will be trained on a dataset of pricing data and will be able to predict the optimal price for new products. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/nsethi31/Kaggle-Predictive-Pricing Difficulty level: Intermediate

  43. Project Name: Disease Diagnosis in Healthcare Description: Develop a model that can diagnose diseases based on patient symptoms and medical test results. The model will be trained on a dataset of patient data and will be able to diagnose new patients. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/nsethi31/Kaggle-Disease-Diagnosis Difficulty level: Advanced

  44. Project Name: Predictive Maintenance in Manufacturing Description: Develop a model that can predict when manufacturing equipment is likely to fail. The model will be trained on a dataset of sensor data from manufacturing equipment and will be able to predict when equipment needs maintenance. Estimated time to complete: 3-4 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/Azure/MachineLearningNotebooks/tree/master/how-to-use-azureml/machine-learning-with-python/predictive-maintenance-modeling Difficulty level: Intermediate

  45. Project Name: Chatbot for Customer Service in Business Description: Develop a model that can simulate human-like conversation for customer service. The model will be trained on a dataset of labeled customer service interactions and will be able to assist new customers. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://github.com/gunthercox/ChatterBot Difficulty level: Intermediate

  46. Project Name: Forecasting Sales in Retail Description: Develop a model that can predict future sales for a retail business. The model will be trained on a dataset of historical sales data and will be able to predict future sales. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/nsethi31/Kaggle-Sales-Forecasting Difficulty level: Beginner

  47. Project Name: Image Analysis for Medical Imaging in Healthcare Description: Develop a model that can analyze medical images for diagnostic purposes. The model will be trained on a dataset of labeled medical images and will be able to analyze new images for diagnosis. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/nsethi31/Kaggle-Medical-Imaging Difficulty level: Advanced

  48. Project Name: Gender Bias Detection in NLP Description: Develop a model that can detect gender bias in natural language processing (NLP) models. The model will be trained on a dataset of labeled text and will be able to detect gender bias in new text data. Estimated time to complete: 2-3 weeks Tech stack used: Python, TensorFlow, Keras, NLTK Online example: https://github.com/tolga-b/debiaswe Difficulty level: Intermediate

  49. Project Name: Fairness in Machine Learning Description: Develop a model that can detect and mitigate bias in machine learning models. The model will be trained on a dataset of labeled data and will be able to detect and mitigate bias in new data. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, Pandas Online example: https://github.com/IBM/AIF360 Difficulty level: Advanced

  50. Project Name: Inclusive Image Captioning Description: Develop a model that can generate inclusive and diverse image captions. The model will be trained on a dataset of labeled images and captions and will be able to generate new captions that are inclusive and diverse. Estimated time to complete: 4-6 weeks Tech stack used: Python, TensorFlow, Keras, OpenCV Online example: https://github.com/google-research/image-captioning Difficulty level: Intermediate

If you found these ideas useful, be sure to subscribe to our blog.


  • mail-2-xxl
  • Twitter

©2023 by Astrania.

bottom of page