![[Neurocle] Logo_white.png](https://static.wixstatic.com/media/242158_0a4304ae861e4107925bc861d88c280f~mv2.png/v1/fill/w_150,h_21,al_c,q_85,usm_0.66_1.00_0.01,enc_auto/%5BNeurocle%5D%20Logo_white.png){kind=small}
Product Overview
Experience the best in each step of your deep learning project
Deep Learning Model Types
9 Types of Deep Learning Models
Neurocle Software provides 9 types(Classification, Patch Classification, Segmentation, Object Detection, Defect Generation : GAN, Anomaly Classification, Anomaly Segmentation, OCR, Rotation) of deep learning vision models.
.png)
Data Management
Efficient and secure
data management
Upload an manage images in various formats. This data is safely stored on your local private network.
Data Management
Tag & Flag for
efficient collaboration
Categorize and manage data conveniently with Tag & Flag. Group images by multicolored tags, and easily spot models and label sets by flags.
%20(1).png)
Preprocessing
Easy data preprocessing
Preprocess your images in many forms without additional software. Crop, rotate, and transform your images however you like.
Labeling
Auto-Labeling
Labels are automatically recommended after the user labels a small number of images.
The initial labeled images set the standard, reducing labeling effort while ensuring accuracy.
Labeling
AI-based Labeling:
Auto-Selector &
Keyword Labeler
By clicking and dragging, our AI detects similar characteristics and identifies related objects.
If you enter simple keywords, the AI will identify and label boxes around the areas in the image that match the keywords.
Model Training
One-click training
for various models
Anyone can create and optimize various deep learning models at the click of a button, even without knowledge of deep learning.
.png)
Model Design & Review
Design and evaluate the model
without programming:
Flowchart & Inference Center
Predict inference results and review your project before deploying the trained models for inference.
Connect multiple models and display the predicted inference results in the GUI. Modify the model design until desired inference time is reached, minimizing trial and error during actual deployment.
%20(1).png)
Report
Clear results through
automatic reporting
Simplify how you view and manage of training results. An automatically generated report displays the training results displayed in organized charts and tables.
Inference
Optimized application
to the platform of choice
From low-spec embedded board to CPUs, GPUs, and NPUs, optimize the trained model for various computing devices and OSs to deploy it on-site.
