Object Detection Cam vs Stream Camera for NDI HX

Object Detection Cam enables you to detect objects in real time using your iPhone or iPad with custom-trained machine learning models. The app includes a built-in model, YOLOv3Tiny.mlmodel, a lightweight neural network optimized for fast object detection across 80 object classes. Object Detection Cam Features: • Load and run custom machine learning models from the built-in library • Rename or delete models directly within the app • Capture photos and save them instantly to the Photos library • Enable sound alerts when an object is detected • Choose from different sound tones for object detection feedback • View all class labels supported by the active machine learning model Adding Custom Models: You can download pre-trained .mlmodel files and add them via AirDrop to the following directory in the Files app: [OD Cam] You may also train your own object detection models using tools such as CreateML or TensorFlow (conversion to .mlmodel required) Custom Model Tips: For instance, if you train a model to detect 5 dog breeds—Husky, Bulldog, Golden Retriever, German Shepherd, and Beagle—it will recognize these accurately. However, detecting a breed like Boxer (not in the model) may yield incorrect results (e.g., identified as Bulldog). To improve accuracy, include as many relevant object classes as possible during training. Usage Examples: • Detect plant species in a jungle environment • Identify dog breeds using a trained model • Detect various animal or insect species • Monitor item status (e.g., identify defective items) • And many more use cases—powered by your creativity Quick Start Guide: 1. AirDrop your .mlmodel file (e.g., DogBreedsModel.mlmodel) into the [OD Cam] folder in the Files app 2. Launch Object Detection Cam 3. Go to the Library tab and tap your custom model 4. The camera will now detect objects based on the model’s class labels 5. When an object is detected, a bounding box and label will appear on screen 6. Tap the shutter button to save a photo to your camera roll [＊] Ensure your model is specifically trained for object detection (not image classification or other model types) What is Object Detection? Object detection uses machine learning or deep learning algorithms to identify and locate objects within images or video. While image classification assigns a label to an entire image, object detection involves localizing one or more objects with bounding boxes and assigning them class labels. Thanks for your support! Visit nitrio.com for more iOS tools and creative apps. Thanks for your support and do visit nitrio.com for more apps for your iOS devices.

Stream Camera for NDI HX allows you to use your iPhone as a video source for your video broadcasting setup. The video is transmitted at up to 4K resolution with fast frame rates. NDI (Network Device Interface) is an IP video protocol that is used for live video streaming and production. It features low latency, making it particularly useful for live productions. This app can be used in an NDI ecosystem of cameras and other devices. The latest NDI 5 standard is supported. Full NDI and NDI HX2 are supported. Features: Video up to 4K Frame rate up to 60 fps Microphone Audio with mute H.264 and H.265 compression for HX2 Full NDI support Camera switching with support for rear wide, front wide, ultra wide, and telephoto cameras Pinch to zoom Zoom and focus control from monitor