Edge AI and Vision Insights: February 26, 2025

 LETTER FROM THE EDITOR

Dear Colleague,

With the 2025 Embedded Vision Summit rapidly approaching (May 20-22 in Santa Clara, California), I invite your company to gain significant visibility and connect with industry leaders by exhibiting and sponsoring at the event. For more details, please see the Prospectus.

Why exhibit and sponsor at the 2025 Embedded Vision Summit?

• Meet the best qualified leads

• • 77% of attendees are developing or planning to develop vision- or AI-based products
  • 96% of attendees would recommend the event

• Interact with focused and influential attendees
  Summit attendees represent a wide range of industries, including agriculture, automotive, consumer electronics, medical, and many more. In 2024, the Summit attracted a highly focused and influential audience:

• • 57% in engineering/engineering management roles
  • 8% in corporate management

• Receive comprehensive sponsorship benefits

• • Access to highly qualified customers and partners
  • Brand visibility through various channels including media partners, LinkedIn, Twitter and newsletters
  • Opportunities to showcase your products and thought leadership

Get in touch with the Summit team to learn more!

Brian Dipert
Editor-In-Chief, Edge AI and Vision Alliance

MULTIMODAL AI IMPLEMENTATIONS

Vision Language Models for Regulatory Compliance, Quality Control and Safety Applications

In this December 2024 Edge AI and Vision Innovation Forum presentation from Carter Maslan, CEO of Camio, you’ll learn how vision language models interpret policy text to enable much more sophisticated understanding of scenes and human behavior compared with current-generation vision models, powering new use cases like ensuring compliance with safety policies and manufacturing regulations.

For more on vision language models, watch the recording of Maslan and colleagues’ recent symposium, “Your Next Computer Vision Model Might be an LLM: Generative AI and the Move From Large Language Models to Vision Language Models“.

How Large Language Models Are Impacting Computer Vision

Large language models (LLMs) are revolutionizing the way we interact with computers and the world around us. However, in order to truly understand the world, LLM-powered agents need to be able to see. Will models in production be multimodal, or will text-only LLMs leverage purpose-built vision models as tools? Where do techniques like multimodal retrieval-augmented generation (RAG) fit in? In this 2024 Embedded Vision Summit talk, Jacob Marks, Senior ML Engineer and Researcher at Voxel51, gives an overview of key LLM-centered projects that are reshaping the field of computer vision and discusses where we are headed in a multimodal world.

AI AND VISION FOR SECURITY APPLICATIONS

Cost-efficient, High-quality AI for Consumer-grade Smart Home Cameras

In this 2024 Embedded Vision Summit presentation, Lin Chen, Chief Scientist at Wyze, explains how his company delivers robust visual AI at ultra-low cost for millions of consumer smart cameras, and how his company is rapidly expanding these capabilities. He begins by introducing Wyze’s edge-AI-enabled consumer cameras and related cloud services, which were created based on goals of delivering “too good to be true” value and building long-term relationships with customers. Chen explains how Wyze has improved model accuracy with user feedback while keeping hardware resource usage to a minimum. Next, he shows how the company has expanded its cameras’ AI capabilities to include things like pet, vehicle and package detection. He also shows how Wyze enables customers to utilize smart cameras as “super-sensors” for home security, replacing conventional single-purpose sensors and overcoming many of their limitations. Finally, he shows how the company is empowering users to add their own personalized AI capabilities, enabling support for numerous diverse use cases.

Implementing AI/Computer Vision for Physical Security Operations

AI-enabled security cameras offer substantial benefits for corporate security and operational efficiency. However, successful deployment requires thoughtful selection of use cases and meticulous attention to managing the technology’s impact on teams. This 2024 Embedded Vision Summit talk explores VMware’s journey through a multiyear digital transformation centered on AI/ML-based security monitoring, decision-making and response. Prasad Saranjame, former Head of Physical Security and Resiliency at VMware, discusses the business, technology and organizational challenges faced, along with best practices for overcoming them. Key strategies include running proof of concept projects, focusing on operational excellence to drive cost reductions and investing in change management to ensure team buy-in. Additionally, he provides insights for others looking to deploy AI in their operations, emphasizing disciplined prioritization, strategic partnerships and a focus on long-term scalability over short-term cost reduction. You will gain actionable insights for effectively harnessing AI in your own organizations.

UPCOMING INDUSTRY EVENTS

Embedded Vision Summit: May 20-22, 2025, Santa Clara, California

More Events

FEATURED NEWS

SiMa.ai Implements DeepSeek-R1 1.5B on Its ONE Platform at Sub-10W Power Consumption

NXP Semiconductors Has Acquired Kinara to Advance Its Edge AI Aspirations and TTTech Auto to Accelerate the Transformation to Software-defined Vehicles

Qualcomm’s Snapdragon X Series Gains an Additional Product Variant, Mini Desktop Form Factors and NPU-powered AI Experiences

Ambarella Partners with Gauzy to Harness the Power of AI for Breakthroughs in Advanced Driver Assistance Systems and LG to Advance AI-driven In-cabin Vehicle Safety Solutions

NVIDIA’s Jetson Orin Nano Developer Kit Receives a 50% Price Cut Along with a “Super” Performance-boosting Software Update

More News