Frequently Asked Questions

Techsolut is a comprehensive computer vision platform that allows you to:

• Annotate images and videos
• Train object detection and classification models
• Deploy models in production
• Integrate computer vision features into your applications
• Manage and analyze your visual data

Our platform is designed for both experts and non-specialists who want to integrate computer vision into their business solutions.

Techsolut can help you solve many computer vision problems, including:

• Object Detection: Identify and locate specific objects in images or videos
• Image Classification: Automatically categorize images
• Segmentation: Precisely delimit object contours
• Text Recognition (OCR): Extract text from images
• Defect Analysis: Detect anomalies or defects in manufactured products
• Object Counting: Automatically count elements in a scene
• Object Tracking: Track moving objects in a video

No, you don't need programming skills to use the main features of Techsolut. Our intuitive user interface allows you to:

• Upload and organize data
• Annotate images using visual tools
• Train models with our training assistant
• Analyze images and videos with trained models
• Visualize results through dashboards

However, for advanced integrations with your existing systems, programming skills may be useful to use our API or SDK. Our documentation provides detailed examples to facilitate this integration.

To create an account on Techsolut:

1. Go to the registration page
2. Fill out the form with your email address, password, and other required information
3. Check your email to confirm your account
4. Log in and complete your profile

You can also sign up using your Google or Microsoft accounts to simplify the process.

Techsolut offers several plans adapted to different needs:

• Free: Limited access to test the platform (5 projects, 100 images, pre-trained models)
• Starter: For small teams and projects (20 projects, 5,000 images, custom training)
• Professional: For teams actively developing computer vision solutions
• Enterprise: For large organizations with specific needs and high volume

Check our pricing page for all the details and to compare the features of each plan.

To modify or cancel your subscription:

1. Log in to your Techsolut account
2. Go to your subscription settings
3. To change plans, click "Change Subscription" and choose the new plan
4. To cancel, click "Cancel Subscription" at the bottom of the page

To get a Techsolut API key:

1. Log in to your account
2. Go to the API Keys section in your account settings
3. Click "Generate New API Key"
4. Give your key a name (e.g., "Web Application", "Testing", etc.)
5. Set usage restrictions if needed (optional)
6. Click "Create Key"

API rate limits vary according to your plan:

• Free: 100 requests/day, max 10 requests/minute
• Starter: 5,000 requests/day, max 60 requests/minute
• Professional: 50,000 requests/day, max 300 requests/minute
• Enterprise: Custom limits based on your needs

API response headers include information about your current usage and limits:

X-RateLimit-Limit: 5000
X-RateLimit-Remaining: 4990
X-RateLimit-Reset: 1619983200

If you reach your limits, the API will return a 429 error (Too Many Requests) and you'll need to wait for your quota to reset.

There are several ways to integrate Techsolut with your application:

1. Via our RESTful API: Use our HTTP API to send images, receive predictions, and manage your resources
2. Using our official SDKs: We offer libraries for Python, JavaScript/Node.js, PHP, Java, and .NET
3. Via Webhooks: Set up webhooks to receive notifications when certain events occur

For quick integration, check our quickstart guide.

For more specific integrations, we recommend:

• For web applications: Our JavaScript SDK or RESTful API
• For mobile applications: Our iOS/Android SDKs or RESTful API
• For background processing: Our Python or Node.js SDK

The training time for a model depends on several factors:

• Amount of data: More images means longer training time
• Model complexity: More sophisticated architectures take longer
• Image resolution: High-resolution images require more resources
• Number of classes/objects: Detecting more types of objects increases complexity
• Your plan: Higher plans have access to more computing power

As a guideline:

• Simple classification model (500 images): 15-30 minutes
• Object detection model (1,000 images): 1-2 hours
• Advanced segmentation model (2,000+ images): 3-8 hours

The number of images needed depends on the complexity of your task:

• Simple classification (2-5 classes): 30-50 images per class
• Complex classification (6+ classes): 50-100 images per class
• Simple object detection: 100-200 annotated images with at least 20-30 examples of each object
• Complex object detection: 300-500 images with objects in various conditions
• Segmentation: 200-500 images with precise contour annotations

Important considerations:

• Images should be varied (angles, lighting, backgrounds, etc.)
• Data should be representative of real-world usage conditions
• A balanced class distribution is important

Here are best practices to improve your model's performance:

1. Improve data quality:
   • Add more diverse images
   • Check annotation quality
   • Balance class distribution
2. Adjust hyperparameters:
   • Modify learning rate
   • Adjust number of epochs
   • Experiment with different batch sizes
3. Try different models:
   • Compare different architectures (YOLOv5, EfficientDet, etc.)
   • Test different model sizes (small, medium, large)
4. Use data augmentation:
   • Enable augmentation options in our assistant
   • Adjust transformation parameters (rotation, zoom, etc.)
5. Analyze prediction errors:
   • Use our error analysis tool
   • Identify common error patterns
   • Add targeted examples for these problematic cases