# qtimlsnpe The qtimlsnpe plugin shows the Qualcomm^®^ Neural Processing SDK capabilities (load and run the models). It accepts a tensor prepared by preprocessing elements such as qtimlvconverter and produces an output tensor that's parsed by postprocessing plugins such as qtimlvclassification, qtimlvdetection, qtimlvsegmentation, and qtimlvpose. This plugin does the following: - Accepts a tensor prepared by the preprocessing elements such as qtimlvconverter. - Produces the output tensor that's parsed by the postprocessing plugins such as qtimlvclassification, qtimlvdetection, qtimlvsegmentation, and qtimlvpose. - Uses the Qualcomm Neural Processing SDK library. - Uses the ION/DMA buffers assigned by the GstMLBufferPool custom buffer pool class through IOCTL commands to the kernel. The model and delegate usage is as follows: - To use a model, provide the absolute path of the model through the model property. For some models, you may have to set the extra layers property. After it's loaded, the model file provides the input and output capabilities: - Number of tensors - Dimensions - Type - If a delegate isn't specified, the plugin executes the model operations on the CPU. - To select a different delegate, set the delegate property accordingly. For more information, see the [Table : Element properties for qtimlsnpe](https://docs.qualcomm.com/doc/80-70029-50/topic/qtimlsnpe.html#section-gf3-q34-myb). GST BIN GstBuffer + GstMemory: Contains a single tensor Example: 1x513x513x3 + GstMemory: Next tensor and so on... + GstVideoMeta: Describes the raw data of a single tensor + GstVideoMeta: Meta for the next tensor entry + ... Qualcomm Open source ML pre-process plugin qtimlsnpe neural-network/tensors neural-network/tensors neural-network/tensors have the following common rules: dimensions: (GstArray) - format is tensor dimension values separated by ',' and marked with delimiters ‘<’ for beginning and ‘>’ for end. # Example-> < <5, 300, 300, 1>, <1, 1, 1000, 1> >##type: (string) {UINT8, INT8, UINT16, INT16, UINT32, INT32, UINT64, INT64, FLOAT32, FLOAT64} First tensor Second tensor Postprocess plugin Sub module **Figure : Tensor mode qtimlsnpe architecture** ## Inheritance chain [GObject](https://docs.gtk.org/gobject/) → [GstObject](https://gstreamer.freedesktop.org/documentation/gstreamer/gstobject.html?gi-language=c) → [GstElement](https://gstreamer.freedesktop.org/documentation/gstreamer/gstelement.html?gi-language=c) → [GstBaseTransform](https://gstreamer.freedesktop.org/documentation/base/gstbasetransform.html?gi-language=c) → GstMLSnpe The following tables provide information on pad templates and element properties of qtimlsnpe. For use cases, see [Qualcomm Neural Processing SDK use cases](https://docs.qualcomm.com/doc/80-70029-50/topic/qualcomm-neural-processing-sdk-use-cases.html). ## Pad configuration | Pad Name | Capabilities | Capabilities | Capabilities | | --- | --- | --- | --- | | SINK template: 'sink'
| neural-network/tensors | type: | { (string)UINT8, (string)INT32, (string)FLOAT32 } | | SRC template: 'src'
| neural-network/tensors | type: | { (string)UINT8, (string)INT32, (string)FLOAT32 } | ## Element configuration Table : Element properties for qtimlsnpe | Property | Description | | --- | --- | | name | The name of the object.
| | parent | The parent of the object.
| | model | Model filename.
| | delegate | Delegate the graph execution to another executor.
| | layers | List of output layers. Should be set if the model has more than one output.
| Last Published: Jun 03, 2026 [Previous Topic qtibatch](https://docs.qualcomm.com/bundle/publicresource/80-70029-50/topics/qtibatch.md) [Next Topic qtimltflite](https://docs.qualcomm.com/bundle/publicresource/80-70029-50/topics/qtimltflite.md) Source: [https://docs.qualcomm.com/doc/80-70029-50/topic/qtimlsnpe.html](https://docs.qualcomm.com/doc/80-70029-50/topic/qtimlsnpe.html)