Realease 2025/6/16 (Target)

Observ June Updates

We’re excited to introduce the June updates to Observ Service, bringing powerful new tools for experimentation, better performance, and broader detection capabilities. Here’s what’s new this month:

New Features:

🧪 LVM Sandbox

The all-new LVM (Real-Time Model) Sandbox is now available! You can upload an image and instantly test detection results using different real-time vision models. This tool helps users explore model behavior, compare outputs, and fine-tune detection prompts—ideal for experimentation and rapid iteration.

🔁 Image-Only Transfer to DataVerse

You can now send individual images to DataVerse without needing associated annotation or detection results. This makes it easier to migrate raw visual data for further processing or labeling.

🧭 New Detection Scenarios: Animals & Weapons

Observ’s detection capabilities have expanded with support for two new built-in scenarios:

• Virtual Fencing for Animal (e.g., for wildlife monitoring or intrusion detection)

• Suspicious Objects & Activities (for security, safety, and threat awareness)

🌙 Night Mode UI

We’ve introduced a brand-new Dark Mode UI, designed for better comfort during late-night monitoring sessions and reduced eye strain in low-light environments.\

🧩 Updated API with New Fields

Our API has been enhanced with new fields and response structures, offering more detailed data and improved integration for developers. Full documentation is available in the updated API reference.

API Configuration

🚀 Streaming Performance Optimizations

More under-the-hood improvements have been made to enhance real-time streaming performance, improving decoding speed, memory usage, and overall system responsiveness—especially under high-load or multi-camera deployments.

---

Coming Soon:

We’re preparing to roll out powerful new features to give users even more flexibility and control over task execution and resource usage:

• Task-Specific Model Selection Soon, you’ll be able to assign a specific model version per task. This allows you to optimize resource usage based on task needs—for example, using lightweight models for lower-resolution cameras or more powerful models for high-precision scenarios.

• New Enhanced Video Task We’re redesigning the video task pipeline:

  • Increased upload limits to handle larger files

  • Batch video upload support under a shared configuration These improvements will significantly improve usability for large-scale or repeated video analysis workflows.

Stay tuned—these upgrades are just around the corner and will further enhance the scalability and efficiency of Observ in real-world deployments.\

---

Support

For technical support, please contact our support team during business hours.