V7Labs

Deep Learning Engineer (Computer Vision)


Remote - London, England, …



The job ad is older than 1 month and may no longer exist.


Job's description


**If you're not familiar with V7, here's what we do: https://youtu.be/iBpgSQk5Qyg

V7 is an AI data platform to automate any visual task, voted by Forbes as one of the top 25 machine learning startups of 2021. We have raised $10 million in venture funding and are backed by the field of AI's greatest leaders, including the minds behind Elixir, Keras, transformers, and more.

We manage the training data and models of hundreds of AI companies and enterprises, spannign tens of thousands of datasets and tens of billions of training data labels. What sets us apart is our team's obsession with pushing our product to where AI will be three years from today.

Here's a bit more on our culture: https://www.v7labs.com/working-at-v7

What you're expected to build with us:

  • General purpose models
  • Instance and semantic segmentation (centroid approaches, high-fidelity semantic segmentation for labelling)
  • Multi-task models (eg. Taskonomy and related papers)
  • Auto-ML in vision
  • Ultra-large model backbones (eg. MOCO and related papers)

We're based on PyTorch, our BE is in Elixir (erlang), and FE is Vue.js. We make GPU machines available both at our office or in the cloud. We are regular CVPR/NeurIPS/ECCV attendees and in normal times attend as a team.

Here are a few things your counterpart(s) worked on in the past 3 months:

  • Auto-Annotate: Automated image segmentation: https://www.youtube.com/watch?v=SvihDSAY4TQ&feature=emb_title
  • An inference and training orchestration engine to run any model request in real-time across a library of computer vision models
  • Several implementations of instance-segmentation architectures like Mask-Rcnn, CenterMask, and object detection architectures like VoVnet
  • Configuration parameters to enable the training of models across various dataset sizes, image sizes, and relative annotation sizes and amounts
  • The development of deep image retrieval systems to automatically rank datasets by content-based parameters without the need of prior training

Requirements

  • Experience in deep learning, from understanding the underlying math, to the ability to illustrate landmark papers within the vision domain (given enough prep time)
  • Evidence of individual contribution to challenging DL/ML projects
  • A CV that demonstrates an early interest in computer science
  • A curious, scientific mind
  • Fluent in English

Benefits

  • Unlimited vacation, just tell us when you need time off
  • Stock options
  • Work from anywhere
  • 7-day company retreats in stunning locations

  • New Apple hardware
  • Paid tickets, accommodation, and travel to relevant conferences, nationally or internationally (NeurIPS, ICCV, CVPR, ...) to expand your network & knowledge during normal times
  • Unlimited high-quality coffee, tea, snacks, and other comforts every day




Apply to position













© Copyright 2022 remotemachinelearning.com. All Rights Reserved.