-
Hi,
Thanks for you works firstly. I have some questions about this network.
First, the slice layers used for padding [ 0 0 1 1] will incerase the time of forward-backward from 60ms to 90ms on my …
-
I used the backbone and some of the layers provided in mmdet to build my custom network, with the backbone using the MobileNetV2 network and the task headers using SElayer to build themselves. The cod…
-
can you shared the test.prototx?thanks
-
I used tensorflow object detection api to train MobileNetV3,https://github.com/tensorflow/models/blob/master/research/object_detection/samples/configs/ssdlite_mobilenet_v3_large_320x320_coco.config, A…
-
请问,除了输入图像大小由300x300提升为320x320,还有其他trick吗?还有为什么这里是90+1个类输出,而不是数据集里标定的80+1个类?
-
Can you write the MobileNetv2-train.prototxt ?
-
I train a model: base cnn part use your mobilenet-ssd; next use SSH M1 M2 M3; change dwconv to DepthwiseConvolution use [https://github.com/yonghenglh6/DepthwiseConvolution](url)
the final model si…
-
大佬!
请问您的动态物体检测方法是什么啊?如何运行呢?
-
Hi there!
Really appreciate your work on explaining CoreML. Super useful!
I cloned your repository to learn about implementing CoreML and to find out how well my SSD_mobilenet_V1 trained on one cla…
-
Hi, chuanqi:
I run the demo to test images, but there occurred a error, `error == cudaSuccess (48 vs. 0) no kernel image is available for execution on the device `
The code was run on jetson t…