We are almost at the end of a new year in Marvik. It was an excellent year, where the company grew a lot and we achieved many goals, but above all a year of personal and professional growth for the whole team, solving challenging technical problems and where we also had a lot of fun … Continued

In this post we attempt to summarize the experience of working on a project that combines signal processing, machine learning and electronics. Its goal was to develop a manipulable sensor-embedded device from which to infer the kind of movement or activity performed by its user. By inference we mean the classification of the activity based … Continued

Are you considering using Machine Learning (ML) in your organization but not sure what for nor how? Or maybe you are just getting around the idea of ML and realized that you don’t know enough about it in order to make an informed decision? If so, this blog post is for you. Disclaimer for techies: … Continued

The following article tries to summarize what is exactly ML Ops and attempts to give a few guidelines for anyone trying to dive into it (or assess if you need to dive into it).  Let me start by saying that MLOps is just the new hype in AI. This has started as a set of … Continued

  In this blog we are going to talk about a technique called model distillation, and how can it be used to train supervised models from synthetic datasets created with GANs. With this technique, we can take advantage of the properties of unconditional image generation and use them in conditional models, with an outstanding improvement … Continued

In this post we will grab a Tensorflow 2 model, optimize it with NVIDIA TensorRT, and deploy it for inference on a JetsonNano. First of all, is optimizing necessary? Short Answer: no. You might have reached to the conclusion that using TensorRT (TRT) was mandatory for running models on the Jetson Nano, this is however, … Continued

Speech synthesis literally means producing artificial human speech. It presents a lot of practical applications, such as music generation, text-to-speech conversion or navigation systems guidance. Traditional approaches for speech synthesis involved searching for speech units that match an input text on a large database and concatenate them to produce an audio file. It often resulted … Continued

At Marvik, we are proud to announce that we were selected to organize an official NeurIPS 2020 meetup. Neural Information Processing Systems annual meeting -aka NeurIPS- is one of the most prestigious Machine Learning conferences. The meetup will be hosted virtually during NeurIPS, from December 7th through December 10th. Stay tuned, more details and confirmed … Continued

Social distancing has been proven to be one of the most effective measures against the spread of the new Coronavirus (COVID19). As part of the so-called ‘New Normality’, many institutions have implemented different ways of controlling distances, from stickers on the floor up to applications or alarm bracelets. This post will present a computer vision … Continued

Quickly validate if SageMaker Neo will lower latency and costs. You have your new shiny model and want to lower latency and costs, let’s get up and running.   In this post we will: Save a trained Keras model Compile it with SageMaker Neo Deploy it to EC2   1. Compile model Where you feed your model to Neo. We will … Continued