# qtimlvdetection The qtimlvdetection plugin processes output tensors of an object detection model from the ML inference plugin (such as qtimltflite, qtimlsnpe, and qtimlqnn) into result of predictions. Note This plugin is going to be deprecated. GST BIN GstBuffer + GstMemory: Contains a single tensor+ GstMemory: Contains next tensor and so on...+ ... + GstVideoMeta: Describes the raw data of a single tensor + GstVideoMeta: Meta for the next tensor entry +... GstBuffer Qualcomm Open source ML pre-process plugin Video source ML inference plugin qtimlpostprocess neural-network/tensors neural-network/tensors video/x-raw qtimetamux text/x-raw **Figure : GstBuffer workflow with qtimlvdetection** ## Postprocessing GST BIN GstBuffer + GstMemory: Contains a single tensor+ GstMemory: Contains next tensor and so on...+ ... + GstVideoMeta: Describes the raw data of a single tensor + GstVideoMeta: Meta for the next tensor entry +... GstBuffer + GstMemory: Image maskmultiple bounding boxes Qualcomm Open source ML pre-process plugin Video source ML inference plugin qtimlpostprocess neural-network/tensors neural-network/tensors video/x-raw qtivcomposer video/x-raw **Figure : Postprocessing for object detection architecture** The negotiated [GstCaps](https://gstreamer.freedesktop.org/documentation/gstreamer/gstcaps.html) determines the processed output. The output can be either of the following: - An image overlay mask (GstCaps: video/x-raw), which are applied over the original image using qtivcomposer. - The element uses the CPU-based [Cairo](https://www.cairographics.org) 2D graphics library to draw the prediction results in ION/DMA buffers. - The GstImageBufferPool custom buffer pool class allocates the ION/DMA buffers through IOCTL commands to the kernel. - The GStreamer-formatted text (GstCaps: text/x-raw) containing the prediction results. The prediction results are parsed into GStreamer-formatted string inside buffers allocated using regular system memory. The module and labels properties of the plugin determine the method for the postprocessing operations. - The module property specifies the postprocessing module. It's populated dynamically at runtime with the libraries available in `/usr/lib/gstreamer-1.0/ml/modules/` containing the prefix "`ml-vdetection`-". - The labels property is a customized text file different for each machine learning detection model that you must provide for the prediction labels. Optional properties are available for adjusting the prediction results. - Use the results property to control the number of results that are displayed - Use the threshold property to set a confidence threshold for prediction. The results with low confidence aren't displayed. ## 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) → GstMLVideoDetection The following tables provide information on pad templates and element properties of qtimlvdetection. For use cases, see the detection use cases in [LiteRT use cases](https://docs.qualcomm.com/doc/80-70029-50/topic/tensorflow-lite-use-cases.html) and [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 | – | – | |
SRC template: 'src'
| video/x-raw | format: | { (string)BGRA, (string)BGRx, (string)BGR16 } | |
SRC template: 'src'
| video/x-raw | format: | { (string)BGRA, (string)BGRx, (string)BGR16 } | |
SRC template: 'src'
| text/x-raw | format: | { (string)utf8 } | | | | | | | | | | | ## Element configuration Table : Element properties for qtimlvdetection | Property | Description | | --- | --- | | name | The name of the object.
| | parent | The parent of the object.
| | qos | Handle the quality-of-service events.
| | module | Module name that's used for processing the tensors.
| | labels | The filename of the labels.
| | results | Number of results to display.
| | threshold | Confidence threshold in %.
| Last Published: Jun 03, 2026 [Previous Topic qtimlaclassification](https://docs.qualcomm.com/bundle/publicresource/80-70029-50/topics/qtimlaclassification.md) [Next Topic qtimlvsegmentation](https://docs.qualcomm.com/bundle/publicresource/80-70029-50/topics/qtimlvsegmentation.md) Source: [https://docs.qualcomm.com/doc/80-70029-50/topic/qtimlvdetection.html](https://docs.qualcomm.com/doc/80-70029-50/topic/qtimlvdetection.html)