Deep Learning Engineer - Speech Recognition
Muse is creating an advanced AI to search the world's video. Today more video is being created than humans can watch.
We are looking for deep learning engineers who want to create the most advanced and intuitive way of searching and discovering within video -- akin recalling a memory in your mind. We are funded by Horizons Ventures, the original backers of Deepmind, Siri, Facebook and Spotify and are serious about creating a world-class team focused on advancing the state of the art in Video understanding.
You will join the machine learning team in sunny Lisbon with a focus on implementing and productionizing the latest machine learning algorithms to analyze video.
Create a next generation speech recognition algorithm in video that fuses signals from visual, sound and speech
recognize and track people across frames of video, and classify the motions they take across scenes.
Create an algorithm that can read the text on objects of video -- so you can find scenes with a BMW 5 series, or where Messi scores a goal.
Interpret queries in natural language that lead to the recall of precise scenes within video.
- Drive the research and implemention of APIs to analyze speech, sound, vision, and motion in video enabling new capabilities in our product
- Drive the research and implementation of systems where humans contribute data to improve these algorithms (e.g reinforcement learning based tagging, clustering-labelling, or evolutionary data capture systems)
- Debug these algorithms across accuracy/computational performance and continue to improve them with a strong cadence of improvements
- Create production ready microservices that can be implemented across our product prototyping algorithms in python and eventually implementing in C/C++
- You can code, you have the right mathematical background, and you have the passion to implement your ideas in production read on
- Research background (PhD/Masters) in video analysis algorithms with emphasis on deep learning OR PhD/Masters in highly quantitative discpline like Astrophysics, Physics, EE -- ability to think from first principles and very high IQ/great attitude acceptable too!
- Experience of training and constructing/tuning deep neural networks in Theano, Tensorflow, PyTorch etc
- Coding fluency in Python and C/C++, and experience of Docker (if not look it up!)
- meticulous detail orientation with data (Garbage in, Garbage out)
- Full-stack scientist - you know the drill of collecting data, cleaning data, analyzing data, prototyping analysis algorithms and profiling your work to iteratively optimize
- You like coding and can show your ideas working -- after all video is visual!
- You manage your time well, and know when to come out of rabbit holes
- You are a great communicator, and positive personality
- You can solve any problem put in front of you
Nice to have
- Software engineering skills outside research for fun - e.g. web/mobile
- You know how to make these algorithms work offline
- Audacious work
- Health benefits
- Food allowances