language

This page is only available in Japanese.

Pose Estimation
Solution

Image illustrating deep learning-based image analysis technology.

Toward a world powered by posture data analysis

Based on pose estimation, motion evaluation and anomaly detection are examined according to use cases. Analysis supports multiple subjects and time-series data, enabling tailored pose estimation use from PoC and validation stages.

Pose Estimation Capabilities

Pose estimation captures posture and movement from video as numerical data. Tracking joint positions, angles, and motion reveals movement characteristics and differences. Beyond recognition, it enables evaluation, comparison, historical tracking, and change analysis for business improvement and research.
Pose Estimation Capabilities

Pose Data for Decisions & Evaluation

  • Moves Over Time

    Moves Over Time

    Using pose data as time-series data enables tracking motion flow and changes. Analyzing movement trends and stability supports evaluation and comparison.

  • Individual Review

    Individual Review

    Pose data can be captured with individual identification, even in video with multiple people or objects. Movement differences can be analyzed for group comparison.

  • Data Integration

    Data Integration

    Pose data can be combined with sensor and numerical data to capture motion and condition, enabling anomaly detection and trend analysis for more advanced decision-making.

  • Challenges Solved with Pose Estimation

    Pose estimation is already being adopted across a wide range of fields. Please feel free to contact us, even at the initial consideration stage.
    • Form Improvement & Health

      Form Improvement & Health

      With a smartphone camera, form issues and posture tilt can be checked instantly, helping improve training and rehabilitation quality.

    • Improve, Standardize

      Improve, Standardize

      Worker movements are analyzed by camera to visualize posture and motion differences, improving training efficiency and productivity.

    • Monitoring & Safety

      Monitoring & Safety

      Continuous camera-based tracking detects falls and abnormal movements without sensors, supporting safety management in care facilities and worksites.

    • Shopping Suggestions

      Shopping Suggestions

      Capturing a standing posture enables analysis of body shape and posture to recommend suitable products, improving the e-commerce experience.

    i-enter Scope

    i-enter supports pose estimation end to end, from issue definition and validation to system integration.

    Hearing & Issues

    Customer goals and operations are reviewed to define pose estimation use and validation scope. Consultation is available even at early stages.

    Validation Planning

    Plans are defined by use case, covering required data, validation points, and PoC or validation phases.

    Estimation Analysis

    Estimation and analysis are designed and validated to meet goals, including multi-subject and time-series analysis.

    Visualization & Link

    Collected pose data and analysis results are clearly visualized and integrated with web and cloud systems for practical business use.

    Results Review & Next Steps

    Validation results are summarized and shared, leading to improvements and the next phase.

    Related Works & Tech

    • AI Development

      AI Development

      Leveraging Deep Learning–based AI Image Recognition, AI solutions are developed for quality inspection and anomaly detection. Recent work includes practical Generative AI to improve processes.

      Learn more
    • Image Analysis System

      Image Analysis System

      An Image Analysis System for Curling that non-invasively analyzes player posture and motion. It captures posture data from video and visualizes movements as 3D models to analyze form differences.

      Learn more
    • Pose Estimation Learning System

      Pose Estimation Learning
      System

      A patented technology (Patent No. 6,868,875) that improves pose estimation accuracy in real video using 3D data, supporting efficient R&D and deployment.

      Learn more
    • Fish Size Estimation Camera

      Fish Size Estimation Camera

      An AI camera for fisheries that measures fish length and height without contact. It detects fish from underwater video to support growth management. A patented technology (Patent No. 6,694,039).

      Learn more
    • AI Fish Counter

      AI Fish Counter

      An automated system that counts fish in aquaculture during grading and shipping. Cameras capture fish and show real-time counts using AI detection, even with overlap (patent pending).

      Learn more
    • Facial Recognition Solution

      Facial Recognition Solution

      Facial Recognition solution for schools, offices, and facilities requiring no dedicated devices. Using smartphones, tablets, or PCs, it enables ID verification and access control.

      Learn more
    • i-enter Sports-Tech Lab

      i-enter Sports-Tech Lab

      We use AI, IoT, and XR to address sports industry challenges. In curling, our systems analyze form with pose estimation, measure sweeping strength, and visualize stone paths. These research-based tools enhance athletic performance with IT.

      Learn more
    • i-ocean

      i-ocean

      We promote Marine Tech for fisheries using AI and IoT. Tools like the Fish Size Estimation Camera for contactless measuring and the IoT Water Quality Sensor for 24/7 monitoring reduce workload and boost quality, supporting smart aquaculture.

      Learn more

    Our Patented Technology

    Fish Size Estimation Device 
    (Patent No. 6694039)

    Patent granted in April 2020

    This patent uses deep learning and stereo cameras to measure swimming fish without contact. It calculates length and height from underwater position data, reducing errors and stress compared to manual methods.
    魚体サイズ算出装置(特許第6694039号)

    Pose Estimation Learning System 
    (Patent No. 6868875)

    Patent granted in April 2021

    This patent is for a system that uses virtual space to build pose estimation models. Previously, manual data creation took time and cost. Now, 360-degree motion is auto-captured with 3D models, and skeletal data is collected at once, speeding up AI development.
    姿勢推定学習システム(特許第6868875号)

    We provide solutions according to your situation and vision

    お客様の状況イメージ画像
    • Discuss technical feasibility.
    • Interested in evaluation from the PoC stage.
    • Consult before data preparation.

    Contact Us

    ご相談、ご質問はお気軽にどうぞ Please feel free to contact us for consultation or questions.

    IoT/AI/XR(VR・AR・MR)/RPA/アプリ開発/Webシステム開発/業務システム開発
    /クラウドインテグレーション/デザイン/UX/デジタルマーケティング関連

    IoT/AI/XR(VR・AR・MR)/RPA/アプリ開発
    /Webシステム開発/業務システム開発
    /クラウドインテグレーション/デザイン
    /UX/デジタルマーケティング関連

    お問い合わせ Contact Us
    TOP