top of page
앵커 1
AI Model
AI Model
Solutions that recognize various characteristics of animals through artificial intelligence
2D Image-based 3D Pose
Estimation analysis model
Deep Learning Model for Estimating Mouse Behavior on a Frame-by-Frame Basis by Detecting Animal Keypoints
2D Image-Based Extraction of
19 Major Body Keypoints and Pose Tracking
Multi-View Video-Based 3D Pose Reconstruction and
Animal Disease-Specific Behavior Analysis System
2D Keypoint
Analysis Model
An AI solution that, when given specific keypoints of an
animal, recognizes those keypoints in images, obtaining
data on a frame-by-frame basis for analysis.
Frame-by-frame, point-by-point Y-axis position
Point-by-point X, Y-axis positions
Paresis Model
AI model recognizes paresised animals
24-hour monitoring allows you to know when paresis occurs and analyze it through stored data.
Monitoring Model
AI model that recognizes the location of each animal part
Data such as animal activity location, area, posture, etc. can be stored and analyzed by recognizing the location of animal parts.
Tumor /
Inflammaion
AI model that recognizes the location and size of an
animal's tumor / inflammation
(Application of thermal imaging camera is required when
checking for inflammation)
Through 24-hour monitoring, the tumor size can be measured to track the size of the tumor and receive an alarm about the
timing of drug administration.
Tail Thickness
Measurement Model
An AI solution that automatically measures parts of
animals' bodies, in this case, digitizes data by recognizing
the tailthickness of mice and the weight printed on paper.
Measurement of tail starting
section (100%) thickness
Measurement of the midsection (50%) thickness of the tail
Measurement of the end section (25%) thickness of the tail
CLVOA OCR
weight recognition
Number, Weight, 100%, 50%, 25% (in pixels)
Vision
analysis Model
Iteratively refining transformer outputs
through expert feedback to ensure high-accuracy cell analysis
Human-in-the-loop Vision Transformer
for cell analysis
AI Cell culture feedback
control model
Classifying cell states (division and death) via deep learning to drive precise 6-DOF robotic micro-manipulation for automated seeding
Automating cell culture analysis
and robotic feedback control
SEM Image analysis
deep learning model
Utilizing transformer-based layers to extract fine-grained topological details from SEM data with high computational efficiency
Precise SEM image segmentation
via Vision Transformer architectures
bottom of page