Launching AI
Customize high-precision AI models through the basic processes and functions of the Launching AI platform
Upload Samples

It supports the creation of datasets in the form of folder structure, and can create training sets, test sets, validation sets, etc. as required. The creation methods include inheriting the existing version and creating a new version. A dataset can create up to 100 dataset versions.


After the dataset version is created successfully, the data can be imported from your local machine or FTP server. The system supports batch upload of image files in the formats of BMP, PNG, JPG and JPEG.

Label
Data processing includes classification and labeling. Image classification supports single-label and multi-label. Image annotation include the methods of points, rectangles, polylines, polygons, line segments and circles, and support various types of annotation tasks. It also supports pre-created annotation templates for multiple points, annotates key points in the target object through a fixed number of key point groups, and the connection relationship between points, automatic average distribution, point attributes, etc. Further improve annotation efficiency with eyedropper and format brush tools.
Train & Evaluate
Select the algorithm type and train the model with the uploaded data. After the model training is completed, the model effect can be verified online. It supports the selection of a variety of algorithms to meet the different needs of performance and effects in different scenarios. It also supports importing your own algorithms, small target detection and other precision optimization functions. Based on our unique pre-training model, a high-precision model with excellent performance can be obtained by training with a small amount of data.
Deploy
After the training is completed, the model can be deployed on a public cloud server or a private server, or packaged into a device-side SDK that can run offline, or purchased a software-hardware integrated solution directly. It can be flexibly adapted to various scenarios and operating environments, and can also be directly released as a device-cloud collaborative deployment package and delivered to edge devices for application.
Vedio Demonstration of Launching AI

Solutions

We provide comprehensive AI data solutions to help enterprises develop their own AI models.
  • Object Detection
    It can automatically classify and count the location, size, and category of objects, create label information through deep learning, identify the target's location, position and posture to confirm the quantity, position, size and posture characteristics of various objects. For example, for the parts that placed out of order, it can realize the gesture recognition and grasping, calculate grasping points, and realize processes such as loading and unloading, sorting and assembly. Through the training and verification process combined with artificial annotation and AI deep learning, the production time, cost and quality can be greatly saved.
    Features:

    Quick verification of detection results online

    Simultaneous detection of multiple targets

    Pixel-level accurate recognition

    Manual annotation combined with deep learning to eliminate the influence of online environment

    Application

    Target Detection

    Position Confirming

    Stack identification

    Position correction

    Existence judgment

    Quantity statistics

  • Segmentation
    Accurate identification and detection of marked feature defects based on deep learning. Upload images for defect feature labeling, and use deep learning technology to train the model to achieve directional identification and detection of labeled feature defects. For example, detecting defects such as cracks, spots, scratches, stain, chromatic aberration, and defects in the images.
    Features:

    Quick verification of detection results online

    Supports simultaneous detection of multiple defects

    Pixel-level detection and custom thresholds

    Multi-scenario general algorithm library available for use

    Application:

    Crack Detection

    Spot Detection

    Scratch Detection

    Stain Detection

    Chromatic Aberration Detection

    Defect Detection

  • Keypoint
    Keypoint detection includes face keypoints, body keypoints, and keypoints of specific types of objects. It is the pre-task of many other algorithms, such as facial expression recognition, human action recognition, etc. Use deep learning and view visibility properties to automatically create keypoints at desired locations, and provide 2D datasets for 3D postures to identify the orientation, position and motion of the instance.
    Features:

    Quick verification of detection results online

    Simultaneous detection of multiple targets

    Key point position from all angles

    Adapt to the situations of occlusion and truncation

    Application：

    Industrial quality inspection

    Sports and Fitness

    Entertainment

    Security

  • Classification
    Automatically clusters massive image datasets to their purpose using deep learning. Quality assurance can be acquired quickly by having labelers validate the AI's output and comparing the labeler's and AI's work.
    Features:

    High precision

    Fast recognition

    High stability

    Full traceability of information

    Application:

    Object Sorting

    Part Defect Classification

    Product distribution and quantity statistics

    Automation of model and equipment program management

  • Data Annotation
    The enterprise-level AI data annotation platform, which is the industry-leading human-machine collaborative AI data labeling platform, integrates data collection and data annotation, and helps enterprises build their AI products. Powerful set of annotation tools can meet complex annotation requirements, such as point, rectangle, polyline, polygon, line segment and circle annotation tools. Rich functions and flexible configuration support intelligent annotation through existing models, further improve the annotation efficiency.
    Features:

    Simple operation

    Rich features

    Smart Annotation with Existing Models

    Custom workflow, multiple rounds of quality inspection to control the quality of data

  • Data Collection
    Collect all kinds of research data required to meet the needs of large-scale data collection and customized data collection. Provide various kinds of raw data according to the customer's data collection scenario requirements, and quickly respond to multi-type data sample collection tasks. Collect data to train the appropriate model based on the customer's use case and target market.
    Features:

    Wide range

    Large scale

    Customization

    Quick response

Industries
Request A Demo
  • First Name*
  • Last Name*
  • Phone Number*

    Telephone error, please check!

  • Email*

    Email error, please check!

  • Company Name*
  • Message*
Submit

After you submitting the message, we will contact you within 3 business days. Thank you!