Deep Learning Tech Talks/Idea Pitching for Lets Stop Wildfires Hackathon

To join online: https://primetime.bluejeans.com/a2m/live-event/yyhzvqpk

Session 1: Deep Learning Tech Talks

Tech Talk 1: Albumentations: fast and flexible image augmentations

Come hear from Kaggle Grandmaster Vladimir Iglovikov https://www.kaggle.com/iglovikov to present about "Albumentations: fast and flexible image augmentations".

Kaggle competitions won with this library.
https://github.com/albu/albumentations/blob/master/docs/hall_of_fame.rst

Vladimir is currently a Senior Computer Vision Engineer at Lyft's Level5 Engineering Center https://level5.lyft.com/. He holds both a Master’s and Ph.D. in physics.

LinkedIn profile: https://www.linkedin.com/in/iglovikov/

Here is the list of Deep Learning challenges won by Vladimir and his team.

1st place: MICCAI 2017: Gastrointestinal Image ANAlysis (GIANA)
1st place: MICCAI 2017: Robotic Instrument Segmentation
1st place: Kaggle: Carvana Image Masking Challenge
2nd place: Safe passage: Detecting and classifying vehicles in aerial imagery
2nd place: CVPR 2018 Deepglobe. Road Extraction

Speaker Bio:
Vladimir got his Ph.D. in Theoretical Condensed Matter Physics at UC Davis. After graduation he was developing Energy Disaggregation algorithms that were a combination of the signal processing and machine learning techniques, working as a data scientist at Bidgely. After this, he moved to San Francisco to work in TrueAccord where he was mainly focused on building recommender systems. Currently, Vladimir is applying Deep Learning techniques to the computer vision problems at the Lyft's Level5 Engineering Centre that is focussed on the development of the self-driving cars.

Tech Talk 2: FUEGO Early Fire Detection

Kinshuk Govil from Team FUEGO (Fire Urgency Estimator in Geosynchronous Orbit) https://fuego.ssl.berkeley.edu/ to present about

"FUEGO Early Fire Detection"

Speaker Bio:
Kinshuk Govil received a BS in EECS from UC Berkeley, and a PhD in Computer Science from Stanford University.

Then he worked at VMware as Principal Engineer, and led the development of vSphere and vCenter products.

Kinshuk developed a machine learning based image recognition system for Benetech [https://benetech.org/] to help digitize books for the blind.
Later, he worked as Principal Engineer at Google, where he worked on a novel SaaS application platform and led ML projects for emerging products.

Currently, Kinshuk is leading the development of ML based early fire detection system with FUEGO group.

Tech Talk 3: Deep Learning Based Large-scale Histological Tau Protein Mapping for Neuroimaging Biomarker Validation in Alzheimer's Disease

Maryana Alegro is currently a computer vision and imaging researcher at UCSF.

Speaker Bio:
Computer Scientist Maryana Alegro is a former BIDS-BCHIS Data Science Fellow, now an Associate Professional Researcher at UCSF. At UC Berkeley, she was a post-doc at the UCSF Grinberg Lab, where she investigated and created computational tools to assist researchers in studying human brain and dementia, especially Alzheimer’s disease (AD). Such tools incorporate a combination of machine learning techniques with visualization and computer vision.

Session 2: Idea Pitching

Pitch your ideas around
1) How to early detect wildfires.
2) How to generate your own electricity and be totally off grid.
3) How PG&E can do a better job rather than shutting off power.
4) How to modernize fire fighting techniques.

Let's Stop Wildfires Hackathon Series was officially launched by AI For Mankind last Wednesday on Oct 16. It will run through Feb 23, 2020.

Hackathon info: https://aiformankind.github.io/lets-stop-wildfires-hackathon/

To participate: https://forms.gle/Rtedxr1rT9HHJogm6

Mentors:
Anna Bethke, Head of AI for Social Good for Intel
Jianming Zhang, Senior Research Scientist from Adobe Research
Jigar Doshi, Machine Learning Lead from CrowdAI
Kinshuk Govil, Lead for Machine Learning based early wildfire detection system with FUEGO
Tim Ball, President of Fireball International