MobileNetV5 is a vision encoder used within Google's multimodal Gemma 3n family of models. It's a lightweight convolutional neural network specifically designed for on-device or embedded applications, prioritizing high efficiency, low latency, and low power usage.
The MobileNet series is known for its use of depthwise separable convolutions to significantly reduce the number of parameters and computational cost compared to traditional convolutional neural networks. This makes them ideal for tasks like image classification, object detection, and semantic segmentation on devices with limited resources, such as mobile phones.

Link to weights: https://huggingface.co/timm/mobilenetv5_300m.gemma3n