At CES 2026, Kickers.ai — together with a professional audio lab — unveiled an AI dual-camera live rig that reconstructs both vision and sound on-device for professional creation and real-time streaming. It is not a “camera plus microphone” bundle but a complete edge imaging-and-audio system redesigned around real creation and live workflows.
Conventional live setups stack a camera, microphone, stabilizer and capture card, raising the setup barrier and pulling a creator’s attention toward operating equipment. Handheld, mobile streaming shakes and tracks slowly, so the subject drifts off-center; a single view cannot balance close-ups and wide shots; ambient noise and lost vocal detail make audio the weakest link; and color, depth of field and style depend heavily on skill and post-production, so beginners rarely get a stable, good-looking result in real time.
Kickers.ai answers with one integrated, one-hand device. A high-compute edge platform sustains a clear, smooth, stable picture under heavy live load. Dual cameras switch focal lengths freely for close-ups, product shots and scene changes. Three-axis mechanical stabilization, combined with self-developed tracking, keeps the subject centered rather than merely steadying the frame. AI imaging lowers the bar for professional output: gestures trigger photo, record and go-live; AI color styles bring the picture to a broadcast-ready state in one tap; and AI cinematic bokeh separates subject from background for natural depth of field. A multi-mic array, backed by the audio lab’s technology, captures omnidirectionally while suppressing noise and enhancing voice. One-tap go-live, fast camera and picture-mode switching, and multi-platform simulcast complete the workflow.
The result is a real-time capture-stabilize-render-stream loop running entirely on-device, turning live streaming from gear assembly into a stable, reusable creative capability — transferable across professional streaming, mobile creation, outdoor shooting and industry content production. This is the meaning of VISION FOR ALL.
Solution: One Integrated Edge AV System
Kickers.ai built the rig as one integrated system rather than an accessory chain, following a closed loop from joint co-creation and AV-system definition to integrated hardware design, on-device algorithm adaptation, same-platform AV integration, live field testing and mass production. Dual-camera imaging, three-axis stabilization, AI gesture, color and bokeh, and multi-mic capture with denoising all run on-device, free of complex peripherals and cloud post-processing.
The dual-camera module and the stabilized gimbal are fused into one structure, and on-device load control keeps every algorithm running in real time within the one-hand form factor. The multi-mic array’s voice processing is backed by a professional audio lab. Facing coupled uncertainties in compute, optics, stabilization pose and acoustics, the team converged the route through rapid validation in Kickers.ai’s POC Studio before formal development. The whole system links one-tap go-live, camera and picture-mode switching, fast parameter tuning and multi-platform simulcast, with full-chain OTA for continuous evolution.
Fig 2: System architecture: Capture → on-device edge AI → workflow & output, fully on-device with OTA
Core Capabilities
- Imaging and stabilization: a high-compute edge platform sustains clear, smooth output under heavy load; dual cameras enable free multi-focal switching; three-axis stabilization with self-developed tracking keeps the subject centered.
- AI imaging: gesture recognition triggers photo, record and go-live; AI color styles and pro-grade filters reach a broadcast-ready look in one tap; AI cinematic bokeh delivers subject-aware depth of field.
- All-scene audio: an omnidirectional multi-mic array suppresses ambient noise and enhances vocal detail, keeping sound clear and natural indoors, outdoors and on the move.
- Workflow and form: one-tap go-live and pause, camera and picture-mode switching, fast parameter tuning and multi-platform simulcast — in a compact, ergonomic, one-hand body.
Fig 3: Dual-lens close-up: the binocular camera head atop the stabilized gimbal
Engineer's Perspective: Technical Decision Stories
The hard part was not choosing features but making them coexist in real time on a one-hand body. A few engineering decisions shaped the result.
- ▍ Compute vs. real-time: rather than chase a higher-compute chip and re-engineer the body and power budget, the team kept the high-compute edge platform and invested in models and scheduling — quantization, parameter pruning and on-device scheduling traded controlled cost for real-time stability, paced “stabilization and tracking first, bokeh and style second.”
- ▍ Compact body, one hand, thermals: a small body with limited power and heat budget still had to carry dual cameras, three-axis stabilization and multi-algorithm AV processing. Fusing the dual-camera module with the gimbal into one structure, plus on-device load control, kept everything real-time and stable with no throttling over long streams.
- ▍ Stability and clean sound in motion: handheld movement causes shake and defocus while complex scenes add noise and echo. On imaging, mechanical stabilization with self-developed tracking and subject-size judgment keeps the frame on the subject; on audio, the multi-mic array with a professional audio lab’s denoising and voice enhancement pushes noise down and lifts vocal detail.
- ▍ Beyond capture: imaging and audio were integrated on one platform with one-tap go-live, mode switching, fast tuning and multi-platform simulcast, and the route was converged through rapid validation in Kickers.ai’s POC Studio with full-chain OTA — ready on-site, evolving remotely.
Outcomes & Industry Value
The project verified that a high-compute edge platform can carry dual cameras, three-axis stabilization, AI imaging and multi-mic pickup simultaneously. It realized an on-device, real-time capture-to-stream loop, upgrading streaming from assembling gear to a stable, reusable creative capability and freeing creators to focus on content rather than equipment.
Four reusable capabilities — edge AV co-design, stabilization and tracking, AI imaging, and low-latency multi-platform streaming — were consolidated and are transferable to professional streaming, mobile creation, outdoor shooting and industry content production. Under stringent constraints of compactness, compute, power and real-time performance, Kickers.ai shows that AI audiovisual capability can be the stable foundation of a professional creation and streaming terminal rather than a stack of gear — making reliable, real and scalable vision and sound within reach for every creator, in the spirit of VISION FOR ALL.