GstTensorIF controls its src-pad based on the values (other/tensor(s)) of its sink-pad.
For example, you may skip frames if there is no object detected with high confidence.
The input/output stream data type is either other/tensor or other/tensors.
-
compared_value: Compared Value (CV), operand 1 (from input tensor(s))
- A_VALUE: Decide based on a single scalar value
- TENSOR_TOTAL_VALUE: Decide based on a total (sum) value of a specific tensor
- ALL_TENSORS_TOTAL_VALUE: Decide based on a total (sum) value of tensors or a specific tensor
- TENSOR_AVERAGE_VALUE: Decide based on a average value of a specific tensor
- ALL_TENSORS_AVERAGE_VALUE: Decide based on a average value of tensors or a specific tensor
-
compared_value_option: An element of the nth tensor
- [C][W][H][B],n (e.g., [0][0][0][0],0 means [0][0][0][0] value of first tensor)
- nth tensor,nth tensor,...
-
supplied_values: Supplied Value (SV), operand 2 (from the properties)
- SV
- SV1,SV2 (for RANGE operators)
-
supplied_values_option: Supplied Value Option
- type of the supplied value (e.g., type:uint8)
-
operator: Comparison Operator
- EQ: Check if CV == SV
- NEQ: Check if CV != SV
- GT: Check if CV > SV
- GE: Check if CV >= SV
- LT: Check if CV < SV
- LE: Check if CV <= SV
- RANGE_INCLUSIVE: Check if SV1 <= CV and CV <= SV2
- RANGE_EXCLUSIVE: Check if SV1 < CV and CV < SV2
- NOT_IN_RANGE_INCLUSIVE: Check if CV < SV1 or SV2 < CV
- NOT_IN_RANGE_EXCLUSIVE: Check if CV <= SV1 or SV2 <= CV
-
then: Action if it is TRUE
- PASSTHROUGH: The element will not make changes to the buffers, buffers are pushed straight through.
- SKIP: Do not generate output frame (frame skip)
- FILL_ZERO: Fill output frame with zeros
- FILL_VALUES: Fill output frame with a user given value
- FILL_WITH_FILE: Fill output frame with a user given file (a raw data of tensor/tensors). If the filesize is smaller, the reset is filled with 0
- FILL_WITH_FILE_RPT: Fill output frame with a user given file (a raw data of tensor/tensors). If the filesize is smally, the file is repeatedly used
- REPEAT_PREVIOUS_FRAME: Resend the previous output frame. If this is the first, send ZERO values.
- TENSORPICK: Choose nth tensor (or tensors) among tensors
[ tensor 0 ]
[ tensor 1 ] -> tensor if -> [ tensor 0 ]
[ tensor 2 ] (tensorpick 0,2) [ tensor 2 ]
input tensors output tensors
-
then_option: Option for TRUE Action
- nth tensor, nth tensor, ... (for TENSORPICK option)
- ${path_to_file}
-
else: Action if it is FALSE
- PASSTHROUGH: The element will not make changes to the buffers, buffers are pushed straight through.
- SKIP: Do not generate output frame (frame skip)
- FILL_ZERO: Fill output frame with zeros
- FILL_VALUES: Fill output frame with a user given value
- FILL_WITH_FILE: Fill output frame with a user given file (a raw data of tensor/tensors). If the filesize is smaller, the reset is filled with 0
- FILL_WITH_FILE_RPT: Fill output frame with a user given file (a raw data of tensor/tensors). If the filesize is smally, the file is repeatedly used
- REPEAT_PREVIOUS_FRAME: Resend the previous output frame. If this is the first, send ZERO values.
- TENSORPICK: Choose nth tensor (or tensors) among tensors
-
else_option: Option for FALSE Action
- nth tensor,nth tensor,... (for TENSORPICK option)
- ${path_to_file}
The format of statement with tensor-if is: If (Compared_Value OPERATOR Supplied_Value(s))) then THEN else ELSE
- Compared_Value and Supplied_Value are the operands.
- Compared_Value is a value from input tensor(s).
- Supplied_Value is a value from tensor-if properties.
If the given if-condition is simple enough (e.g., if a specific element is between a given range in a tensor frame), it can be expressed as:
gst-launch ... (some tensor stream) !
tensor_if name=tif \
compared-value=A_VALUE compared-value_option=3:4:2:5,0 \
operator=RANGE_INCLUSIVE \
supplied-values=10,100 \
then=PASSTHROUGH \
else=TENSORPICK \
else-option=0 \
! tif.src_0 ! (tensor(s) stream for TRUE action) ...
! tif.src_1 ! (tensor(s) stream for FALSE action) ...
However, if the if-condition is complex and cannot be expressed with tensor-if expressions, you may create a corresponding custom filter with tensor-filter, whose output is other/tensors with an additional tensor that is "1:1:1:1, uint8", which is 1 (true) or 0 (false) as the first tensor of other/tensors and the input tensor/tensors.
Then, you can create a pipeline as follows:
gst-launch ... (some tensor stream)
! tensor_filter framework=custom name=your_code.so
! tensor_if name=tif \
compared_value=A_VALUE \
compared_value_option=0:0:0:0,0 \# 1st tensor's [0][0][0][0].
operator=EQ \
supplied_values=1 \
then=PASSTHROUGH \# or whatsoever you want
else=SKIP \# or whatsoever you want
! tif.src_0 ! tensor_demux name=d \ #for TRUE action
d.src_0 ! queue ! fakesink # throw away the 1/0 value.
d.src_1 ! queue ! do whatever you want here...
! tif.src_1 ! (tensor(s) stream for FALSE action) ...
gst-launch ... (some tensor stream) !
! tensor_if name=custom_if
! (tensor_stream)
...
register call back function for custom_if
int TensorIfCallback(const GstBuffer * in_buf, GstBuffer * out_buf) {
/* Define condition and handle buffer */
}
gst-launch-1.0 v4l2src ! videoconvert ! videoscale ! video/x-raw, format=RGB, width=640, height=480, framerate=30/1 ! tensor_converter ! mux.sink_0 \
videotestsrc ! videoconvert ! videoscale ! video/x-raw, format=RGB, width=640, height=480, framerate=30/1 ! tensor_converter ! mux.sink_1 \
tensor_mux name=mux ! tensor_if name=tif \
compared_value=A_VALUE \
compared-value-option=0:0:0:0,0 \
supplied-value=70,150 \
operator=RANGE_INCLUSIVE \
then = PASSTHROUGH\
else = TENSORPICK \
else-option= 1 \
tif.src_0 ! tensor_demux name=demux \
demux.src_0 ! tensor_decoder mode=direct_video ! videoconvert ! ximagesink async=false sync=false \
demux.src_1 ! tensor_decoder mode=direct_video ! videoconvert ! ximagesink async=false sync=false \
tif.src_1 ! tensor_decoder mode=direct_video ! videoconvert ! ximagesink async=false sync=false