Qualcomm® AI HubAI Hub
HomeCompute ModelsLaMa-Dilated

LaMa-Dilated

High resolution image in-painting on-device.

LaMa-Dilated is a machine learning model that allows to erase and in-paint part of given input image.

Snapdragon® X Elite
TorchScripttoQualcomm® AI Engine Direct
91.9ms
Inference Time
4MB
Memory Usage
333NPU
Layers

Technical Details

Model checkpoint:Dilated CelebAHQ
Input resolution:512x512
Number of parameters:45.6M
Model size:174 MB

Applicable Scenarios

  • Image editing

Licenses

Source Model:APACHE-2.0
Deployable Model:AI Model Hub License

Tags

  • backbone
    A “backbone” model is designed to extract task-agnostic representations from specific data modalities (e.g., images, text, speech). This representation can then be fine-tuned for specialized tasks.

Supported Compute Chipsets

  • Snapdragon® X Elite