NOON47

脸肯定是大于20*20的，一种图片在尺度遍历的时候： while (minWH >= 12) { scales.push_back(scale); minWH *= factor; scale *= factor; } total_boxes_.clear(); uint8_t *pImg = get_img(img); for (int i = 0; i < scales.size(); i++) { int...

谢谢提醒，ZQ_CNN_MTCNN已经调通的了，比较才发现有问题，现在想起来，这个代码用的是之前的参数，所以检测不出来，现在经过对应ZQ_CNN_MTCNN的参数修改，可以了，只是框抖动还是很大，谢谢左博

@ghimiredhikura hi, > hi.. I can convert [ssr2_megaage_1_1/model-0000.params, ssr2_megaage_1_1/model-symbol.json] model to ncnn format. : [ssr2_megaage_1_1_ncnn.zip](https://github.com/wayen820/gender_age_estimation_mxnet/files/4756525/ssr2_megaage_1_1_ncnn.zip) > > Now, the problem is at inference time, the predicted output is just garbage...

this problem have been solved!!!! the ncnn code must be like this: ncnn::Mat im_info(1,3); im_info[0] = 0; im_info[1] = 1; im_info[2] = 2; ex.input("data", in); ex.input("stage_num0", im_info); ex.input("stage_num1", im_info); ex.input("stage_num2",...