Visium

03/10/2022

EPFL Students, we’re looking forward to meeting you.

We’ll be awaiting you at our stand (S21) during the EPFL Start-up Day on the 4th of October.

Eléonore Homasson, Fady Sader and other team members will be here to meet you.

See you tomorrow ✌️

PS: Can you solve the challenge?

27/05/2021

Our COO, Matteo Togninalli has been awarded the ETH Medal, recognizing his PhD thesis as the best dissertation within his department at ETH Zürich🏅

Matteo's thesis focused on exploring the genotype-phenotype relationship using big data and Machine Learning.

If you are interested in digging deeper into his thesis, you can find it linked below 👇

Congratulations Matteo! We are thrilled and extremely proud of your achievement and to have you as one of the leading forces at Visium.🚀

https://www.research-collection.ethz.ch/handle/20.500.11850/448681

20/04/2021

Researchers at UC Berkeley introduce Bottleneck transformers: a modified convolutional architecture with Transformer blocks for visual recognition🔍

This model, acronymed BoTNet, combines different, yet complimentary, strengths of self-attention and convolutions to achieve incredible performance outcomes. In image classification tasks, the model achieves 84.7% top-1 accuracy on the ImageNet benchmark, and does so 2.33 times faster in compute time than the popular EfficientNet models.🚀

Check out this week’s digest to find out how the model achieves these impressive results, and the implications of its application in industry!

Other topics in this week’s issue:
✔️Sparse attention Transformers for long text processing
✔️Seven legal questions that every data scientist should ask when building Machine Learning solutions.

Researchers at UC Berkeley introduce Bottleneck Transformers: a modified convolutional architecture with Transformer blocks for visual recognition.

01/04/2021

We are happy to welcome back William CappellettiCappelletti as a Junior ML Engineer!🚀

William completed his higher education at , obtaining a BSc in Mathematics and MSc in Applied Mathematics with a minor in Computer Science. While at EPFL, William was a teaching assistant, and a member and class delegate of the AGEPoly Committee. In 2019, William joined for an internship as a Junior Machine Learning Engineer, before broadening his professional portfolio as a research scientist for Quantum Machine Learning at Entropica Labs, Singapore. At Entropica, William designed and developed Machine Learning algorithms for Quantum Computers and published a paper at IEE International Conference on Quantum Computing and Engineering. William then expanded and cleaned his code to create a ready-to-use Python library to apply the algorithm to actual hardware or simulators.

Ultimately William’s desire to work on a variety of different projects, using numerous techniques, and apply cutting-edge technologies practically motivated him to join our team again!

Williams' passion, curiosity, and ambitious nature is clearly manifested in his work, and we are excited to have his contribution on our projects!

Read Williams publication here: https://lnkd.in/dTMYjxY

01/04/2021

We are happy to welcome two new Jr. ML Engineers - Camilla Casamento Tumeo and Pauline Maury Laribiere 🎉

Camilla is joining straight from her MSc in Data Science from ETH Zürich, which she completed after obtaining a BSc in Mathematics from the Università degli Studi di Padova. Despite being early in her professional career, Camilla's self-motivated nature and extracurricular experiences are impressive. Camilla is a multi-time gold medalist of the Italian Mathematical Olympiads, and 2-time bronze medalist of the European Girls' Mathematical Olympiads. During her studies Camilla worked on multiple ML projects, including psychiatric disorder classification from behavioral and MRI data, and human 3D pose estimation.

Pauline completed her higher education at , obtaining a BSc in Microengineering, and MSc in Microengineering majoring in Robotics and minoring in Biocomputing. Prior to joining , Pauline worked as a ML Engineer at Sensat in London, where she worked on Computer Vision projects with large scale point clouds and orthomosaics. Pauline also wrote her Master Thesis at Sensat, on Deep Learning Semantic and Instance Segmentation for 3D Clouds and 2D Orthomosaics at a city scale.

Welcome, we are very excited to have you on board!

01/02/2021

While the current situation continues to challenge day-to-day activities, our team continues to grow persistently!🚀

We are welcoming a new Junior Machine Learning Engineer to the team - Hugo Porta.

Hugo has an M.Eng in Computer Science and Applied Science from Télécom Paris, the leading French School in Computer Science, and a MSc in Data Science from the Université Paris-Saclay, one of France’s top-ranking Master’s of science in Machine Learning. To complement his impressive academic background, Hugo gained extensive experience as a Data Scientists at Credit Suisse. During his time at Credit Suisse, Hugo prototyped and developed an industry-leading unauthorized trading detection model and trader risk monitoring pipeline.

Ultimately, his desire to investigate a wide range of ML theories and data types, and work on a diverse portfolio of projects, motivated Hugo to join .

We are extremely excited to have Hugo on board, and are confident that his insights will contribute greatly to our projects and solutions!

22/01/2021

Staying true to our commitment to the community, we are excited to announce Visium's decision to track, reduce and ultimately eliminate our carbon footprint 👣♻️

As part of this decision, we have audited our 2019 carbon footprint, with our partner myclimate , and compensated our emissions. We are excited to announce that we operate net-carbon-neutral since 2019.

This is just the first step in our journey to eliminating our carbon footprint. We are now launching our carbon audit for 2020 and continue striving to reduce our footprint here on out.

If you want to learn more about our sustainability efforts you can check out our blog post below!

As our first step towards our broader sustainability strategy, we are officially a certified carbon neutral company since 2019.

21/01/2021

A warm welcome to our newest Machine Learning Engineer, Elli Tzini!🚀

Elli has a Bachelor in Chemistry from the National & Kapodistrian University of Athens and a MSc in Computational Biology and Bioinformatics from the ETH Zürich, where she was also awarded a certificate of Advanced Studies in Computer Science. To complement her impressive educational background, Elli's passion for Machine Learning and Data Science pushed her to explore different areas in the industrial sector. She has a broad range of work experience as a Machine Learning Engineer, including in EM Imaging, Behavioural Sciences and Legal Tech.

Ultimately her eager desire to work alongside active learners and to continue developing her career in Machine Learning in a meaningful way motivated her interest in Visium.

We are very excited to see Elli's impressive expertise and talents materialize in our projects!

Very happy to have you on the team! 🙌

11/01/2021

Ever wondered what a Kangaroo made of chili looks like? Maybe not, but now you can see it! 🦘🌶️

Microsoft-backed research institution OpenAI shows impressive progress in text-to-image synthesis.

Learn about how the model, called DALL-E, combines GPT-3 with state-of-the-art Computer Vision techniques to generate images out of a text prompt. The only limit is your imagination!

Read about this, privacy considerations in Large Language Models, and deep learning tumour contouring, in our first digest of 2021!

https://lnkd.in/dsyv83N

Privacy considerations in large language models. Text to image synthesis. AI employed to detect tumours and speed up cancer treatments.

23/12/2020

As 2020 comes to an end, we reflect on the moments we shared with all of you! Thank you all for your trust, support, and loyalty.
The Visium family wishes you happy holidays and a great start to the new year. We have great things planned for 2021 and look forward to sharing them with you.
Stay safe and healthy,
Your Visium Team

07/12/2020

DeepMind's AlphaFold solves the 50-year-old problem of protein folding decades before experts had predicted🧬💡

Read more about how they achieved this major scientific breakthrough in this week's digest!

We also discuss the addition of 'Too Long;Didn't Read' in Scientific literature and Reducing Robot Bias.

DeepMing's AI solution Predicts Protein Folding. Too Long To Read in Scientific Literature. Making Robots Less Biased than Humans.

02/12/2020

Robots are coming for your trash🤖 ♻️

AMP Robotics, a recycling robotics technology developer which has raised over $23 million in venture financing, is changing the future of recycling.

Learn how the technology achieves 99% accuracy on their complex object detection task, and get other curate AI news in the link below!

Recycling AI make waste management financially viable. Facebook uses ML to identify harmful content. Ethical considerations of GANs.