AI Engineer

Results 6 issues of AI Engineer

左博你好! 有没有使用过 mtcnn mxnet模型转化为onnx格式啊? export_model 函数要求一个input_shape,PNet的input_shape是变化的,这个参数怎么设置啊? ` import onnx import mxnet as mx import numpy as np from mxnet.contrib import onnx as onnx_mxnet import logging logging.basicConfig(level=logging.INFO) from onnx import checker...

你好,我使用下面的代码测试 人脸识别的工程,模型数据你提供的, 发现相同的人score反而低,不同人反而高?测试的图片是使用 data目录下是图片(00.jpg,00_.jpg, 010.jpg ) ![00](https://user-images.githubusercontent.com/4198916/57579988-cd6ff280-74d6-11e9-9f80-2952fa49960a.jpg) ![00_](https://user-images.githubusercontent.com/4198916/57579989-d52f9700-74d6-11e9-8a7c-392de76128e7.jpg) ![010](https://user-images.githubusercontent.com/4198916/57579990-d95bb480-74d6-11e9-854e-a44bc0e31c13.jpg) `int main() { #if ZQ_CNN_USE_BLAS_GEMM openblas_set_num_threads(1); #elif ZQ_CNN_USE_MKL_GEMM mkl_set_num_threads(1); #endif ZQ_FaceRecognizer* recognizer[1] = { 0 }; std::string prototxt_file = "model/model-r50-am.zqparams"; std::string...

Hi, the linked http://insightface.ai/files/models/arcface_r100_v1.zip and http://insightface.ai/files/models/retinaface_r50_v1.zip all are failed, Can I download these models from other link? thinks