转换的3个脚本如下：
1、0_import_model
NAME=yolov3
ACUITY_PATH=…/bin/
convert_caffe=${ACUITY_PATH}convertcaffe
convert_tf=${ACUITY_PATH}convertensorflow
convert_tflite=${ACUITY_PATH}convertflite
convert_darknet=${ACUITY_PATH}convertdarknet
convert_onnx=${ACUITY_PATH}convertonnx
$convert_tf
–tf-pb ./model/yolov3_dx_18003__27_3anchors_608.pb
–inputs input/input_data
–input-size-list ‘608,608,3’
–outputs conv_sbbox/BiasAdd\ conv_mbbox/BiasAdd\ conv_lbbox/BiasAdd
–net-output ${NAME}.json
–data-output ${NAME}.data
2、1_quantize_model
NAME=yolov3
ACUITY_PATH=…/bin/
tensorzone=${ACUITY_PATH}tensorzonex
#dynamic_fixed_point-i8 asymmetric_affine-u8
$tensorzone
–action quantization
–dtype float32
–source text
–source-file ./data/validation_tf.txt
–channel-mean-value ‘0 0 0 255.0’
–model-input ${NAME}.json
–model-data ${NAME}.data
–model-quantize ${NAME}.quantize
–quantized-dtype dynamic_fixed_point-i8
–quantized-rebuild
3、2_export_case_code
NAME=yolov3
ACUITY_PATH=…/bin/
export_ovxlib=${ACUITY_PATH}ovxgenerator
$export_ovxlib
–model-input ${NAME}.json
–data-input ${NAME}.data
–reorder-channel ‘2 1 0’
–channel-mean-value ‘0 0 0 255.0’
–export-dtype quantized
–model-quantize ${NAME}.quantize
–optimize VIPNANOQI_PID0X88
–viv-sdk ${ACUITY_PATH}vcmdtools
–pack-nbg-unify
首先第一个问题，./data/validation.txt 中./path/img.jpg, 813。这个813是什么意思？比如我们模型输入是608×608×3,这里813要改成608吗？
第二个问题，–reorder-channel ‘2 1 0’ ，按照我们上面模型输入608×608×3，这里是0 1 2 ，还是2 1 0,我试了两个好像最后运行时候都会报错，报错信息如下：
model.width:3
model.height:608
model.channel:608
E detect_api:[Inputsize not match! net VS img is height:608vs608, width:channel:608vs3]
感谢感谢！@Frank
