Erwann Millon

CTO

San Francisco, California, United States7 yrs 2 mos experience

Most Likely To SwitchAI Enabled

Key Highlights

Led the development of cutting-edge AI tools for creatives.
Trained the world's largest open-source realtime video model.
Published research cited by major tech labs.

Stackforce AI infers this person is a leading expert in AI-driven creative tools and generative models.

Contact

Skills

Core Skills

AiDeep LearningComputer VisionGenerative Ai

Other Skills

EnglishFrenchKubernetesMLOpsPythonPyTorchMachine LearningAmazon Web Services (AWS)Data ScienceArtificial Intelligence (AI)TorchNumPyDockerBashPyTorch Lightning

About

ML researcher / engineer specialized in image/video gen AI Turning cutting-edge research into powerful tools for creatives

Experience

7 yrs 2 mos

Total Experience

2 yrs 6 mos

Average Tenure

3 yrs 4 mos

Current Experience

Krea.ai

Head of Applied AI

Feb 2023 – Present · 3 yrs 4 mos · San Francisco Bay Area · On-site

led training of Krea 2 Editing model
designed novel 0-shot style reference system for Krea 1 and 2, allowing t2i with interpolable style conditioning from reference images. beats frontier models including Google's
co-led research on Krea 2 from pretraining to post-training,
co-led post-training on Krea 1 and flux1.krea, SOTA image models tuned for aesthetics
published and presented at major conferences (SIGGRAPH, ICML), cited by labs like Nvidia, Tencent, Huawei
trained Krea Realtime 14B, world's largest open-source realtime video model
built world's fastest inference for realtime t2i and image editing with LoRA swapping
designed fast and versatile image model finetuning pipeline allowing users to train LoRAs on custom datasets in under 2 minutes
built and scaled inference and infra on k8s to millions of users

AIDeep Learning

Self-employed

Machine Learning Engineer

Jan 2021 – Feb 2023 · 2 yrs 1 mo

Latent Face Editor
Built full Web UI for face editing
Implemented adversarial gradient-masked loss to perform localized edits.
Wrote custom backpropagation hooks in the PyTorch autograd engine
Color Diffusion
World's first diffusion model to tackle image colorization
UNet architecture using L channel of LAB color space as conditioning, and diffusion over AB channels
FashionGen
Built a multi-model Al pipeline (Bounding boxes w/ segmentation models -> BLIP image-to-caption -> Stable diffusion finetuning)
Built an interface allowing users to generate and novel outfits
Indexed CLIP embeddings of fashion images for text
and visual semantic search of fashion items
Multi-GPU training on AWS EC2
Virtual Clothes Try-On (Dreambooth)
Used 10-20 images of individual clothing articles to finetune Stable Diffusion allowing users to apply a real outfit onto a picture of themselves.
Built automated pipeline to scrape images from a product page, process them, and launch distributed training
Used textual-inversion and prompt engineering to optimize results
VQGAN-CLIP reimplementation
Simple implementation focused on readability and ease-of-use: 3 lines of code to generate or edit an image
Featured as a research project on the Huggingface transformers repo

Computer VisionGenerative AI

Careceiver

Co-Founder

Jul 2019 – Jul 2021 · 2 yrs · Lisbon Area, Portugal

Startup founded during European Innovation Academy
Built an MVP for a SaaS that helps caregivers care for chronically ill relatives
Incubated at HAG startup accelerator

Freelance

Translator (French/English) and Voice Over Artist

Apr 2019 – Feb 2022 · 2 yrs 10 mos

Led large-scale multi-month projects
Specialized in e-learning, clients include the Government of Canada, UNICEF, UPL
Experience from translation to performing to audio engineering

EnglishFrench