TDA4VM: 使用TAD4部署时， 报如下错误，Error pulling tensor from GST Pipeline。我发现原因为param.yaml中，crop为(800,288),resize为(800,288),如果将其改为整数则不报该错。resize必须是一个整数吗？

==========[INPUT PIPELINE(S)]==========

[ERROR] Error pulling tensor from GST Pipeline

[PIPE-0]

multifilesrc location=/opt/edgeai-test-data/videos/video_0000_h264.h264 loop=True ! h264parse ! v4l2h264dec ! tiovxmemalloc pool-size=8 ! capsfilter caps="video/x-raw, format=(string)NV12;" ! tiovxmultiscaler name=split_01 split_01. ! queue ! capsfilter caps="video/x-raw, width=(int)1280, height=(int)720;" ! tiovxdlcolorconvert out-pool-size=4 ! capsfilter caps="video/x-raw, format=(string)RGB;" ! appsink max-buffers=2 drop=True name=sen_0 split_01. ! queue ! capsfilter caps="video/x-raw, width=(int)784, height=(int)760;" ! tiovxmultiscaler target=1 name=tiovxmultiscaler2

param.yaml参数如下：

postprocess:
data_layout: NCHW
preprocess:
crop:
- 800
- 288
data_layout: NCHW
interpolation: null
pad_color:
- 114
- 114
- 114
resize:
- 800
- 288
reverse_channels: false
session:
artifacts_folder: 'artifacts'
input_mean:
- 0.406
- 0.456
- 0.485
input_optimization: false
input_scale:
- 0.225
- 0.224
- 0.229
input_shape:
images:
- 1
- 3
- 800
- 288
model_folder: model
model_path: model/culane_18.onnx
session_name: onnxrt
target_device: pc
task_type: detecion