Lensmera Depth Dav2 Small is an ONNX model designed for on-device depth-based background blur, commonly known as bokeh. It is a verbatim mirror of the onnx-community/depth-anything-v2-small model, specifically renamed for use in the Lensmera app. The model supports dynamic input shapes and is provided in the fp16 ONNX format with opset 14. According to the licensing information, only the Small variant is released under the Apache-2.0 license, while other variants such as Base, Large, and Giant are licensed differently and are not permitted for the same use. The model has not been modified from its upstream version aside from renaming, and it is intended for application in scenarios where depth-based background blur is required directly on the device. No information is provided about deployment as an inference provider or integration with other platforms. The model is available for download from Hugging Face, but download statistics are not tracked. Further technical or usage details are not specified in the available evidence.
In the Other AI space, Lensmera Depth Dav2 Small takes a focused approach. Enabling developers to add depth-based background blur to images on-device without cloud processing. Lensmera Depth Dav2 Small is an open-source project aimed at mobile developers, computer vision researchers. The project is open source (Apache-2.0). The product ships for the web.
It is developed by Lensmera, and the product first shipped in 2024. The project is developed in the open on GitHub with 8.4k stars. Among its 4 catalogued features are depth estimation, background blur, and ONNX format.
Latest indexed changes and source events
lensmera/lensmera-depth-dav2-small verified by the PulseGate indexer
Other apps tracked under the same category.
