Thoughts and Trends in Tech

Discover the DGA's impact on data sharing, privacy, and innovation, and how ML6 solutions align with its objectives.

Unlock Machine Learning with Amazon SageMaker JumpStart: A Comprehensive Guide

Fine-tune large language models for custom behavior and knowledge, weighing open source vs. closed source options.

At the end of college years, it can be very confusing to start determining your next steps. Are you 100% convinced about your choice of study? Do you have doubts about how your education plays out in practice? These questions certainly don't come as a surprise. There is a lot coming at you. For this reason, it is very interesting to explore the field in practice. Luckily, schools often offer that option through internships. 

Amazon recently announced their foundation model service: Amazon Bedrock. We got our hands on it as part of a limited preview. So, what’s in it for you?

Unlocking ASR Excellence: Fine-Tuning Whisper for Dutch Language

AI-Enhanced Product Photography: Revolutionizing E-Commerce Visuals

Understanding this behaviour and explaining why full control is unachievable

With the growing maturity of AI models and techniques, the technical art of the possible is no longer the most important ceiling.Company specific boundaries will come from the strategy of a company and the values it adheres to.

The magical world of ChatGPT, troves of information, and levels of automation.

Addressing ethical considerations in LLM development.

Visualising Dynamic Pricing modelling to build insights

Securing AI deployment in OT environments.

Discover the potential of Google's PaLM 2 as we harness generative language models for achieving business success.

Learn how dynamic pricing has revolutionized e-commerce through probabilistic statistics, iterative modeling, and customer feedback.

Is AI only applicable within factories and large corporate organizations? Anything but. AI will (and already is) creating impact within the Human Resources branch as well. We speak with Julie Plusquin, Talent Partner at ML6 about what the evolution within HR may look like.

Learn how to make your home battery smarter by connecting it to external inputs such as energy prices and weather. This can save up to 35% on electricity costs.

Harness large foundation models affordably with LoRA technique for business applications.

ML6 tackles enzyme stability prediction in Kaggle competition, pushing boundaries in life sciences.

AI image generation without copyright infringement using a Creative Commons dataset.

Comparing GCP and Azure from a developer's perspective.

Different methods available for fine-tuning Large Language Models (LLMs) and when to use them.

Introduction to Foundation Models and choosing the right one.

The structure of FMOps to optimize your foundation model workflows.

Unleash Astonishing Animations with Stability AI's Stable Animation SDK

Google just released its Generative AI offering. What is it and what does it mean for you?

Recording of the webinar on Generative AI for corporate use: How to get started with Large Language Models

Generative AI as your co-pilot: specialized AI models and their impact on design innovation.

Discover 3D optical acquisition methods: Stereo Vision, Structured Light, Time of Flight, and LiDAR.

Discover the power of 3D computer vision and its impact on industries. Learn about advantages, applications, and trends.

Exploring Recent Developments in Open Source LLMs and Their Implications.

In this blogpost we explore the different options out there around usage of ChatGPT and other GPT models. For every option we we determine the risks around compliance, security and privacy.

In this article, we delve into the mind of Kaat Van Doren, Head of Talent & Culture at ML6, as she explores the role of curiosity within the organization and her own professional life at a podcast sit down with Cevora.

LoRA is revolutionizing fine-tuning for large AI models, enabling efficient adaptation without straining resources. Explore its principles, effectiveness, and impact on the open-source community.

Discover key amendments proposed in the EU AI Act, including banned and high-risk applications, regulations for foundation models and generative AI, and the impact on open-source components. Prepare for upcoming AI regulations and their implications.

Explore the potential of large language models (LLMs) for businesses: models, tooling, challenges, and applications

This blog post explores the process of labelling speech data for Automatic Speech Recognition (ASR).

We'll cover the basics of language models and how they can be trained on your organization's data to improve search accuracy, automate tagging, and even generate new content

Jens Bontinck, Head of Delivery & Advice, shares his view on the status quo of AutoGPT. In this blogpost, you will not learn how to run it, nor will I go into details of the technical solution. Therefore, if you are a product owner, a functional analist, or a commercial enthousiast, this should be right up your alley. Happy to read your feedback!

Discover the truth behind the hype of ChatGPT & GPT-4. Read about the potential game-changing capabilities of GPT-4 or if it's just another gimmick.

Jens Bontinck shares his view on a recent technique called Reflexion

How we empower our agents in their own personal road to success.

In this blogpost we cover the most important aspects of tackling a computer vision problem by labeling your own data.

Improving tennis court line detection using machine learning for near real-time predictions.

A guide on how to choose the right speaker diarization framework for your use case.

Tips on figuring out how to navigate the recruitment process and determine which companies are a good match for you.

This blog post explores the current state of AutoML tools and their ability to deliver on the promise of automating machine learning tasks without expert knowledge.

Apache Beam has lots of I/O connectors available, but not all connectors are in your language of choice.

Key take-aways of Google Cloud’s new Visual Inspection AI tool.

Explore the practical implementation of dynamic pricing, its effectiveness, challenges, and the use of Thompson sampling to optimize prices.

Examples of AI use cases for climate action and clean energy and sustainable industry and agriculture.

Examples of AI use cases for transport & mobility and environment & oceans.

Discover how AI is driving sustainability in energy systems and manufacturing, with pragmatic use cases for decarbonization and waste reduction.

Log4Shell got security teams around the world on their toes In this blog post, we take a look at how and why we built a live threatmap on top of Google Cloud to detect and visualize cyber attacks and log4j exploits.

Identified gaps in the architectures for SLR/SLT by considering and exploiting existing state-of-the-art architectures.

Semantic search from a business point of view and how it can bring value to your business processes.

This blogpost covers the need for MLOps and how it helps to manage the complexity of machine learning models and deployments. Secondly, it addresses the business implications of implementing MLOps as a proper ecosystem. Lastly, the challenges of building an MLOps capability are covered to ensure the success of your machine learning initiatives at your organisation.

Deploying a Scikit-learn Pipeline + LightGBM Classifier Model on Vertex AI for Explainable Online Predictions

Explore the field of Hybrid AI, combining machine learning and expert rules for optimal outcomes. Learn its relevance, integration methods, and real-world examples.

Common hybrid NLP design patterns and look at example situations of when you should opt for which pattern.

How to approach uncertainty in machine learning : forecasting, regression and classification models.

When deploying a Transformer model, the complexity increases since besides the model one also needs to consider how to deploy the tokenizer. Is it best to deploy the tokenizer on the server or to have the end users handle it and what is the best way to deploy the tokenizer along with the model on the server so that inference latency and costs are minimized?

What do we do though if we don’t have any or enough usable data yet to get started? One obvious option is to collect (more) data. However, today we want to go beyond that. We will look at three other dimensions that can be relevant to unblock ML use cases: data protection, external data, and synthetic data.

Applications of using VQGAN combined with Discrete Absorbing Diffusion models.

We review and compare the previous SOTA generative models, from autoencoders to autoregressive models, and work our way to the discrete absorbing diffusion model.

Developing a DALL-E-based logo generator.

Detecting errors in transformers generated abstractive summaries

How we tackle information management at ML6 and how our tech radar helps us do this.

In this blogpost we will discover the complexity of pooling that hides behind its apparent simplicity.

Learn more about the similarities and differences between transformers and sentence-transformers.

Learn how to improve training data quality and maximize AI model accuracy in our webinar on data-centric AI, with practical tips and advanced approaches.

A practical guide to generating artificial data using Unity Perception.

Insights in how to run your Transformer in the cloud in a cost-effective way.

How to maximise the value of data to drive better performance of ML models while minimising the cost of doing so.

Assessing Detic for object detection of thousands of classes.

In this blogpost we’ll set up a docker container to run an NVIDIA deepstream pipeline on a GPU or a Jetson and show you some tricks to get the most out of it! Deepstream’s documentation, guides and sample projects are few and far between, so this article aims to be a reference to get you from Deepstream zero to Deepstream-hey-I-can-make -a-pipeline-now and beyond!

Abbreviated version of the AI act proposal : what you need to know.

Open Sourcing Dutch Summarization Datasets for Transfer Learning

A practical overview on information retrieval, in particular Semantic Search.

Why we created an Ethical AI unit at ML6 and what we learned so far.

In this blogpost we take a look at what data augmentation is and how you can efficiently discover good data augmentation strategies for your model and your use case.

A short introduction to unsupervised anomaly detection and how to apply it.

In light of a partnership with the organizations Reef Support and Fruitpunch AI, we designed an image segmentation algorithm. The main challenge here was the way the data is labelled, containing sparse pixel annotations created by domain experts. The blogpost covers on a high level how we solved this challenge, and how the challenge will be going forward.

In this blogpost, we look at object detection using 3D ‘depth’ sensor data. We explain how two recent approaches, VoteNet and 3DETR, tackle the problem and benchmark their performance out of the box and after fine-tuning.

Discover the importance of Explainable AI (XAI) in understanding machine learning models and explore practical case studies, including Google's Vertex Explainable AI tool.

Train neural networks faster with JAX on TPUs

Introduction blogpost to JAX : how it works, how it is different from Tensorflow/Pytorch and why we think every machine learning engineer should add it to its arsenal of tools.

We explore how the connexion library facilitates API-first design with Python andtake a look at what the advantages are of API-first and compare with the opposite approach of “code-first”. Based on a benchmarking exercise and our values, we will dive deeper into why we are helping to maintain connexion.

We turn the wheel and make the most unproductive week, the most productive week of the year with our XMas Projects.

Deepstream 6.0 Python Boilerplate for Custom AI-based Multi-Sensor Processing and Video Understanding

Learn about three ways that knowledge graphs and machine learning reinforce each other.

Pretrained TabTransformer performs well in unsupervised pretraining on tabular data, but LightGBM outperforms it on small labeled datasets.

Learn how to use the ByT5 model for OCR correction, a token-free model that operates on raw bytes of text.

Businesses in industries such as media, finance, jurisdiction and many more, tend to accumulate a large quantity of medium to long documents. Processing these documents can be challenging. We'll deep dive into the three main trends here.

A practical guide to efficient neural networks: model compression techniques including pruning, quantization, knowledge distillation, and optimization tricks like gradient checkpointing and accumulation.

Learn how to structure your data with knowledge graphs in this introduction blogpost to knowledge graphs and its business applications.

Realtime Tennis Action Recognition with Movinet Stream: Sports Video Analysis in the Real World

We are often asked how to detect very small objects in high resolution, i.e. very large images. A good example is finding objects in aerial images. The goal of this blog post is to demonstrate a practical approach to this problem using Slicing Aided Hyper Inference (SAHI).

In this blogpost, our ML engineer Juta will dive into the world of multimodal AI. Multimodal AI is able to reason about different modalities, such as text and images, at the same time and leads to more general understanding across domains.

Overcome manufacturing challenges with AI and Google Cloud. Learn from practical ML6 use cases and leverage machine learning retraining programs for improved operations.