![[2602.17537] IRIS: Learning-Driven Task-Specific Cinema Robot Arm for Visuomotor Motion Control](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.17537] IRIS: Learning-Driven Task-Specific Cinema Robot Arm for Visuomotor Motion Control
Summary
The paper presents IRIS, a cost-effective robotic arm designed for cinematic motion control using learning-driven techniques, achieving high precision and ease of use.
Why It Matters
IRIS addresses the challenges of high costs and complexity in robotic camera systems, making advanced cinematography more accessible. Its innovative use of imitation learning and affordability could democratize filmmaking and enhance creative possibilities in various industries.
Key Takeaways
- IRIS is a 6-DOF manipulator designed for autonomous cinematic motion control.
- The system utilizes a lightweight, fully 3D-printed design, reducing costs to under $1,000.
- It employs a goal-conditioned visuomotor imitation learning framework for smooth camera trajectories.
- Real-world tests show high accuracy in trajectory tracking and reliable performance.
- IRIS can generalize across diverse cinematic motions, enhancing its versatility.
Computer Science > Robotics arXiv:2602.17537 (cs) [Submitted on 19 Feb 2026] Title:IRIS: Learning-Driven Task-Specific Cinema Robot Arm for Visuomotor Motion Control Authors:Qilong Cheng, Matthew Mackay, Ali Bereyhi View a PDF of the paper titled IRIS: Learning-Driven Task-Specific Cinema Robot Arm for Visuomotor Motion Control, by Qilong Cheng and 2 other authors View PDF HTML (experimental) Abstract:Robotic camera systems enable dynamic, repeatable motion beyond human capabilities, yet their adoption remains limited by the high cost and operational complexity of industrial-grade platforms. We present the Intelligent Robotic Imaging System (IRIS), a task-specific 6-DOF manipulator designed for autonomous, learning-driven cinematic motion control. IRIS integrates a lightweight, fully 3D-printed hardware design with a goal-conditioned visuomotor imitation learning framework based on Action Chunking with Transformers (ACT). The system learns object-aware and perceptually smooth camera trajectories directly from human demonstrations, eliminating the need for explicit geometric programming. The complete platform costs under $1,000 USD, supports a 1.5 kg payload, and achieves approximately 1 mm repeatability. Real-world experiments demonstrate accurate trajectory tracking, reliable autonomous execution, and generalization across diverse cinematic motions. Subjects: Robotics (cs.RO); Machine Learning (cs.LG) Cite as: arXiv:2602.17537 [cs.RO] (or arXiv:2602.17537v1 [cs.RO] for ...