Efim Boieru
AI & Computer Vision Expert | Tech Leader | Deepfake Detection
Get in Touch
Building Advanced Computer Vision at Scale
I’m a computer vision and AI leader with 10+ years of experience turning research into production systems across startups and global tech companies. At Incode, I lead a growing team of 7 ML/CV engineers and researchers building passive face liveness and deepfake detection to secure digital identity without adding user friction.
I enjoy building strong teams, setting technical direction, and shipping systems that stay reliable as threats evolve.
Current Focus at Incode
I lead the development of passive face liveness and deepfake detection for identity verification. The work is driven by generalization to new attacks, low-friction UX, and production constraints like latency and device variability.
Face Liveness Detection
Multimodal, video-first liveness that combines RGB + depth + temporal cues to reduce presentation-attack risk without active user challenges.
Deepfake Prevention
Generalization-focused deepfake detection evaluated against 100+ generators and robust to common transformations (compression, resizing, post-processing) that appear in real verification flows.
Real-Time Processing
Production models optimized for <1 second end-to-end latency, with monitoring and dataset refresh loops to keep performance stable as threats evolve.
Recent Highlights
Purdue University Benchmark
Our face deepfake detection was evaluated in a Purdue study and ranked among the top-performing solutions in their comparison.
Deepsight Product
Liveness and deepfake detection are core capabilities of Incode’s Deepsight product launch.
Industry Impact & Collaboration
Huawei
Developed machine learning algorithms for telecommunications and mobile technology, enhancing performance across global networks.
Bosch
Created advanced machine learning systems for automotive and industrial applications, pushing boundaries in safety and automation.
MIT
Collaborated on cutting-edge research projects, contributing mathematical rigor and innovative approaches to complex optimization challenges.
Incode
Leading face liveness and deepfake detection initiatives, developing next-generation authentication systems for enterprise security.
Speaking & Knowledge Sharing
I regularly speak at international conferences and technical meetups, where I share insights on the latest advancements in AI security, computer vision, and machine learning applications.
I aim to bridge the gap between research and real-world implementation, so listeners leave with ideas they can actually explore and apply in their own work.
International AI Conferences
Talks on face liveness and deepfake detection and biometric security trends shaping the future of authentication.
Technical Meetups
Advanced challenges in deepfake detection, multimodal AI, real-time performance, and achieving zero-error targets in production systems.
Academic Collaborations
Research partnerships exploring theoretical foundations and emerging methodologies in computer vision and AI.
Selected Talks

22:31

YouTube

Multimodal Video Face Liveness: A Better Alternative? | Efim Boieru | DSC EUROPE 24

Efim explored multimodal video face liveness as a more robust alternative to traditional face liveness detection techniques. He examined the limitations of single-frame passive and video-based active methods, highlighting failure cases. Efim explained how combining depth estimation, motion analysis, and temporal data in multimodal video liveness offers a more comprehensive approach to facial authentication. He discussed the use of foundational models and the integration of multiple modalities li

53:31

YouTube

Madrid Machine Learning Meetup: Face anti-spoofing and deepfake detection

We are thrilled to host Efim Boeru, a leading expert in AI, computer vision, and mathematics, with over eight years of experience developing advanced machine learning solutions for industry giants like Huawei, Bosch, and MIT. Currently driving innovation in face liveness and deepfake detection at Incode, a US-based unicorn startup, Efim combines technical expertise with a passion for problem-solving and cutting-edge research. In this enlightening session, Efim will delve into multimodal video

51:40

YouTube

3D Human Avatars: Efim Boeru (Twin3D, MIPT)

Наш первый гость в 2022 году — Ефим Боеру, CTO в компании Twin3D, выпускник МФТИ и Сколтеха. Ефим расскажет про процесс создания 3D сканов людей, их аватаров и цифровых двойников. Мы подробно обсудим, какие именно алгоритмы ха этим стоят, и какие предъявляются аппаратные требования к установкам для многокамерной съемки. Вступайте в наш Telegram: https://t.me/arhub_channel, чат https://t.me/+VS5bli0JhmE5ZTEy и группу ВК: https://vk.com/arhub_ai, там уютно, дружелюбно и интеллектуально 🤓 Ссылк

Core Expertise Areas
Deep Learning
Architecting and training neural networks from basic CNNs to advanced architectures including transformers, vision-language models (VLMs), and attention mechanisms for complex visual recognition tasks.
Computer Vision
Human-centric computer vision tasks including deepfake detection, face anti-spoofing, generative AI applications, 3D human reconstruction, and facial recognition systems.
Biometric Security
Creating robust authentication solutions that balance security with user experience, specializing in liveness detection and anti-spoofing techniques.
Applied Mathematics
Leveraging mathematical foundations to optimize algorithms, improve model performance, and solve challenging computational problems.
Multimodal Models
Fusion architectures combining multiple data modalities with advanced neural networks for real-time inference pipelines in production environments.
AI Innovation
Staying ahead of emerging trends, from generative AI to multimodal learning, and applying them to real-world security challenges.
Technical Achievements
10+
Years Experience
Delivering production-grade AI solutions across industries
70+
ML Models
Developed and deployed at enterprise scale
100M+
Users Protected
Through advanced authentication systems
25+
Presentations
At conferences and technical events worldwide
Let's Connect
If you’re interested in collaboration, speaking opportunities, or exploring new ideas in AI and computer vision, I’m always open to a conversation.
I work with researchers, engineers, and organizations looking to push beyond established approaches — whether through technical collaboration, advisory work, or shared exploration of emerging methods.