Does the neural network support text input or is it limited to images?

Kevin_Guzman

Hello,

I would like to know if it is possible to input text data into my neural network model, or if it is designed exclusively for processing images. Could you provide information on this?

I greatly appreciate your assistance.

Regards.

jakaskerl

Hi Kevin_Guzman
The neural network in the DepthAI platform is not limited to image inputs. You can create a message on the host or Script node, populate the tensor with the data, and send the message to the input of the NeuralNetwork node. This message carries tensors and its data. However, the formatting of the AI model input frame uses BGR color order and Planar / CHW (Channel, Height, Width) layout conventions. If the model accepts color images, it should accept 3 channels (B,G,R), and if it accepts grayscale images, it should accept 1 channel.

It's important to note that any OpenVINO neural networks can be run using this node, as long as the VPU supports all layers. This allows you to pick from 200+ pre-trained model from Open Model Zoo and DepthAI Model Zoo and directly run it on the OAK device.

Thanks,
Jaka