LOOP-i
2024
AI Hardware. Wearable Design. Spatial Interaction.
Project Duration
2 Months · Group Project (Led by Me)
My Roles
Project Management, Interaction Designer, Industrial Designer
Goal
Exploring the potential of AI technologies and new hardwares designed to leverage them
Collaboration
OPPO x Royal College of Art
London Design Festival 2024
OPPO x RCA: Connective Intelligence
With the advancement of AI technologies, I believe the best way to harness its potential is by integrating it into scenarios where it is truly needed.
LOOP-i is an AI-powered wearable camera that reimagines the vlogging experience. It uses multimodal inputs to perceive creators’ intent and capture what they focus on. By leveraging generative AI, it enables creators to generate dynamic content based on their surroundings.
Context
In the past few years vlogging has been trending rapidly, especially travel vlog. Here are some data. Among all vlogs, travel vlog takes around 1/3. Travel vlogs also have higher profit increase and there is a 106% annual growth of travel vlog channels.
LOOP-i
Project Duration
2 Months · Group Project (Led by Me)
My Roles
Project Management, Interaction Designer, Industrial Designer
Goal
Exploring the potential of AI technologies and new hardwares designed to leverage them
Collaboration
OPPO x Royal College of Art
2024
AI Hardware. Wearable Design. Spatial Interaction.
With the advancement of AI technologies, I believe the best way to harness its potential is by integrating it into scenarios where it is truly needed.
LOOP-i is an AI-powered wearable camera that reimagines the vlogging experience. It uses multimodal inputs to perceive creators’ intent and capture what they focus on. By leveraging generative AI, it enables creators to generate dynamic content based on their surroundings.
London Design Festival 2024
OPPO x RCA: Connective Intelligence
Challenge
Travel vlogging has been so popular but the recording devices remains traditional. Here we are targeting at amateur vloggers, but even they don’t have many choices.
Most people take hand-held cameras for recording, which really interrupts their travel experience as their hands are always occupied. There are smaller devices like the body cam but it has fixed perspective and are not that flexible to change dynamically.
What if there is an intelligent companion that truly understands your perspective and unleashes your creative exploration?
Concept
LOOP-i is a pair of AI-powered wearable cameras designed for creators, capturing content from an eye-level perspective. Worn primarily on the ears, it provides a natural, immersive view of what users focus on.
LOOP-i leverages AI in two key ways: First, it uses multimodal inputs—such as images, voice, and hand gestures—to automate content recording and simplify the process. Second, it incorporates generative AI, enabling users to creatively transform and stylize their footage based on their surroundings, unlocking new possibilities for personalized content creation.
Feature 1 - Multimodal Framing Assistance
By combining motion sensing, voice recognition, and gesture recognition, the device intelligently centers on desired targets for hands-free content creation. This experience is further enhanced by pairing LOOP-i with a smartwatch, providing camera previews and additional controls directly on the wrist.
LOOP-i follows head direction by default to capture first person view
Scenario View
Expected Camera View
LOOP-i centers targets in the frame based on creators’ speech and where their fingers are pointing
Scenario View
Expected Camera View
AI’s Visual Inference (Technical Prototyping)
Groq’s AI platform, using a connected webcam for real-time image capture, processes and summarizes visual data, demonstrating the feasibility of AI for real-time context understanding.
Gesture-based Interactions
What you see is what’s captured. A series of classic framing hand poses have been developed for vloggers to perform multiple actions, allowing them to create vlogs naturally.
Feature 2 - Gesture-based Interactions
What you see is what’s captured. A series of classic framing hand poses have been developed for vloggers to perform multiple actions, allowing them to create vlogs naturally.
Crop hand poses provides an intuitive way to quickly frame the view
Gesture-based interaction offers more opportunities to capture key moments
ML Based Gesture Recognition (Technical Prototyping)
Teachable Machine was used for machine learning-based gesture training and recognition. By training on hundreds of hand pose images, it demonstrates high accuracy in recognizing gestures.
Creative Capture
The integrated Gen AI enables creative vlogging. Creators can trigger content generation by simply focusing on a target in the environment and speaking to the device to specify the effect they want.
Feature 3 - Creative Capture
The integrated Gen AI enables creative vlogging. Creators can trigger content generation by simply focusing on a target in the environment and speaking to the device to specify the effect they want.
The combination of crop gestures and Gen AI allows stylised and creative filming creation
Beyond video content generation can be applied to photos and selected targets
Product Details
The style of design language is approachable and futuristic. The body as well as the support is applied with rounded form. The combination of dark glass, camera and LED provides a feeling that the device is alive and understand people.
LOOP-i is also designed to be used in a diverse ways. By default it is designed be to ear-mounted but can also be attached on palm for selfie or placed on a surface to capture 3rd person-view.
Process
Despite a strong vision for applying emerging AI technologies, our approach remained user-centric. We began with thorough user research, engaging vlogger friends and reviewing numerous vlogs to uncover key insights. One consistent finding was that handheld filming devices are often disruptive, and content creators seek more personalized solutions based on real-time capture. These insights guided our focus towards integrating Visual and Generative AI with wearable designs, while testing various form factors to optimize ergonomics and usability.
Next Steps (WIP)
I am currently developing a functional prototype using hardware and digital tools that offer enhanced features and greater flexibility for modification. I’m exploring visually focused platforms such as Raspberry Pi and OpenCV to create more advanced prototypes that can better reflect the intended user experience and align more closely with the original concept targeting at multimodal interaction and auto framing features.
Context
In the past few years vlogging has been trending rapidly, especially travel vlog. Here are some data. Among all vlogs, travel vlog takes around 1/3. Travel vlogs also have higher profit increase and there is a 106% annual growth of travel vlog channels.
Challenge
Travel vlogging has been so popular but the recording devices remains traditional. Here we are targeting at amateur vloggers, but even they don’t have many choices.
Most people take hand-held cameras for recording, which really interrupts their travel experience as their hands are always occupied. There are smaller devices like the body cam but it has fixed perspective and are not that flexible to change dynamically.
What if there is an intelligent companion that truly understands your perspective and unleashes your creative exploration?
Concept
LOOP-i is a pair of AI-powered wearable cameras designed for creators, capturing content from an eye-level perspective. Worn primarily on the ears, it provides a natural, immersive view of what users focus on.
LOOP-i leverages AI in two key ways: First, it uses multimodal inputs—such as images, voice, and hand gestures—to automate content recording and simplify the process. Second, it incorporates generative AI, enabling users to creatively transform and stylize their footage based on their surroundings, unlocking new possibilities for personalized content creation.
Process
Despite a strong vision for applying emerging AI technologies, our approach remained user-centric. We began with thorough user research, engaging vlogger friends and reviewing numerous vlogs to uncover key insights. One consistent finding was that handheld filming devices are often disruptive, and content creators seek more personalized solutions based on real-time capture. These insights guided our focus towards integrating Visual and Generative AI with wearable designs, while testing various form factors to optimize ergonomics and usability.
Multimodal Framing Assistance
By combining motion sensing, voice recognition, and gesture recognition, the device intelligently centers on desired targets for hands-free content creation. This experience is further enhanced by pairing LOOP-i with a smartwatch, providing camera previews and additional controls directly on the wrist.
LOOP-i follows head direction by default to capture first person view
Scenario View
Expected Camera View
LOOP-i centers targets in the frame based on creators’ speech and where their fingers are pointing
Scenario View
Expected Camera View
AI’s Visual Inference (Technical Prototyping)
Groq’s AI platform, using a connected webcam for real-time image capture, processes and summarizes visual data, demonstrating the feasibility of AI for real-time context understanding.
ML Based Gesture Recognition (Technical Prototyping)
Teachable Machine was used for machine learning-based gesture training and recognition. By training on hundreds of hand pose images, it demonstrates high accuracy in recognizing gestures.
Creative Capture
The integrated Gen AI enables creative vlogging. Creators can trigger content generation by simply focusing on a target in the environment and speaking to the device to specify the effect they want.
Gesture-based Interactions
What you see is what’s captured. A series of classic framing hand poses have been developed for vloggers to perform multiple actions, allowing them to create vlogs naturally.
The combination of crop gestures and Gen AI allows stylised and creative filming creation
Crop hand poses provides an intuitive way to quickly frame the view
Product Details
The style of design language is approachable and futuristic. The body as well as the support is applied with rounded form. The combination of dark glass, camera and LED provides a feeling that the device is alive and understand people.
LOOP-i is also designed to be used in a diverse ways. By default it is designed be to ear-mounted but can also be attached on palm for selfie or placed on a surface to capture 3rd person-view.
Gesture-based interaction offers more opportunities to capture key moments
Beyond video content generation can be applied to photos and selected targets
Next Steps (WIP)
I am currently developing a functional prototype using hardware and digital tools that offer enhanced features and greater flexibility for modification. I’m exploring visually focused platforms such as Raspberry Pi and OpenCV to create more advanced prototypes that can better reflect the intended user experience and align more closely with the original concept targeting at multimodal interaction and auto framing features.
LOOP-i
Project Duration
2 Months · Group Project (Led by Me)
My Roles
Project Management, Interaction Designer, Industrial Designer
Goal
Exploring the potential of AI technologies and new hardwares designed to leverage them
Collaboration
OPPO x Royal College of Art
2024
AI Hardware. Wearable Design. Spatial Interaction.
With the advancement of AI technologies, I believe the best way to harness its potential is by integrating it into scenarios where it is truly needed.
LOOP-i is an AI-powered wearable camera that reimagines the vlogging experience. It uses multimodal inputs to perceive creators’ intent and capture what they focus on. By leveraging generative AI, it enables creators to generate dynamic content based on their surroundings.
London Design Festival 2024
OPPO x RCA: Connective Intelligence
Context
In the past few years vlogging has been trending rapidly, especially travel vlog. Here are some data. Among all vlogs, travel vlog takes around 1/3. Travel vlogs also have higher profit increase and there is a 106% annual growth of travel vlog channels.
Challenge
Travel vlogging has been so popular but the recording devices remains traditional. Here we are targeting at amateur vloggers, but even they don’t have many choices.
Most people take hand-held cameras for recording, which really interrupts their travel experience as their hands are always occupied. There are smaller devices like the body cam but it has fixed perspective and are not that flexible to change dynamically.
What if there is an intelligent companion that truly understands your perspective and unleashes your creative exploration?
Concept
LOOP-i is a pair of AI-powered wearable cameras designed for creators, capturing content from an eye-level perspective. Worn primarily on the ears, it provides a natural, immersive view of what users focus on.
LOOP-i leverages AI in two key ways: First, it uses multimodal inputs—such as images, voice, and hand gestures—to automate content recording and simplify the process. Second, it incorporates generative AI, enabling users to creatively transform and stylize their footage based on their surroundings, unlocking new possibilities for personalized content creation.
Multimodal Framing Assistance
By combining motion sensing, voice recognition, and gesture recognition, the device intelligently centers on desired targets for hands-free content creation. This experience is further enhanced by pairing LOOP-i with a smartwatch, providing camera previews and additional controls directly on the wrist.
LOOP-i follows head direction by default to capture first person view
Scenario View
Expected Camera View
LOOP-i centers targets in the frame based on creators’ speech and where their fingers are pointing
Scenario View
Expected Camera View
AI’s Visual Inference (Technical Prototyping)
Groq’s AI platform, using a connected webcam for real-time image capture, processes and summarizes visual data, demonstrating the feasibility of AI for real-time context understanding.
Gesture-based Interactions
What you see is what’s captured. A series of classic framing hand poses have been developed for vloggers to perform multiple actions, allowing them to create vlogs naturally.
Crop hand poses provides an intuitive way to quickly frame the view
Gesture-based interaction offers more opportunities to capture key moments
ML Based Gesture Recognition (Technical Prototyping)
Teachable Machine was used for machine learning-based gesture training and recognition. By training on hundreds of hand pose images, it demonstrates high accuracy in recognizing gestures.
Creative Capture
The integrated Gen AI enables creative vlogging. Creators can trigger content generation by simply focusing on a target in the environment and speaking to the device to specify the effect they want.
The combination of crop gestures and Gen AI allows stylised and creative filming creation
Beyond video content generation can be applied to photos and selected targets
Product Details
The style of design language is approachable and futuristic. The body as well as the support is applied with rounded form. The combination of dark glass, camera and LED provides a feeling that the device is alive and understand people.
LOOP-i is also designed to be used in a diverse ways. By default it is designed be to ear-mounted but can also be attached on palm for selfie or placed on a surface to capture 3rd person-view.
Process
Despite a strong vision for applying emerging AI technologies, our approach remained user-centric. We began with thorough user research, engaging vlogger friends and reviewing numerous vlogs to uncover key insights. One consistent finding was that handheld filming devices are often disruptive, and content creators seek more personalized solutions based on real-time capture. These insights guided our focus towards integrating Visual and Generative AI with wearable designs, while testing various form factors to optimize ergonomics and usability.
Next Steps (WIP)
I am currently developing a functional prototype using hardware and digital tools that offer enhanced features and greater flexibility for modification. I’m exploring visually focused platforms such as Raspberry Pi and OpenCV to create more advanced prototypes that can better reflect the intended user experience and align more closely with the original concept targeting at multimodal interaction and auto framing features.