In the previous article, I reached mAP 0.770 for VOC2007 test. Four months has past. After trying a lot of interesting ideas from different papers, such as FPN, celu, RFBNet, I finally realised that the data is more important than network structures. Then I use COCO2017+VOC instead of only VOC… Read more »
I got a desktop computer to train deep learning model last week. The GPU is GTX1050TI with 4GB memory which is enough for basic training on object detection. But the CPU is too old. Therefore when I run the training process, the idle of CPU is 0%. I need to… Read more »
SSDLite is a variant of Single Shot Multi-box Detection. It uses MobileNetV2 instead of VGG as backbone. Thus it can make detection extremely fast. I was trying to implement SSDLite from the code base of ssd.pytorch. Although it’s not a easy work, I finally learn a lot from the entire… Read more »
1. Type convertion in Numpy Here is my code:
|
1 2 3 4 5 6 |
import numpy as np a = np.asarray([1, 2]) b = [] c = np.concatenate((a, b)) print(c.dtype) |
Guess what? The type of variable ‘c’ is ‘float64’! Seems Numpy automatically considers a empty array of Python as ‘float64’ type. So the correct code should be:
|
1 2 3 4 5 |
import numpy as np a = np.asarray([1, 2]) b = [] c = np.concatenate((a, np.asarray(b, dtype=a.dtype)) |
This time, the type of ‘c’ is ‘int64’ 2. Convert a tensor… Read more »
In the previous article, I reached mAP 0.739 for VOC2007. After about two weeks, I add more tricks to reach mAP 0.740. The most important trick is escalating the expand-scale of augmentation which is made from this patch. Increase the scale range could help the model to detect a smaller… Read more »
Previously, I was using CUB-200 dataset to train my object detection model. But after I used CUB-200-2011 dataset instead, the training loss became ‘nan’.
|
1 2 3 4 |
iter 10 || Loss: 17.9996 || timer: 0.2171 sec. iter 20 || Loss: nan || timer: 0.2145 sec. iter 30 || Loss: nan || timer: 0.2145 sec. ... |
I tried to reduce the learning rate, change optimizer from SGD to Adam, and use different types of initializer for parameters. None of these solved… Read more »
1. ‘()’ may mean tuple or nothing.
|
1 2 |
len(("birds")) # the inner '()' means nothing len(("birds",)) # the inner '()' means tulple because of the comma |
The result is:
|
1 2 |
5 1 |
2. Unlike TensorFlow’s static graph, PyTorch could run neural network just as the code. This means a lot of conveniences. The first advantage, we could print out any tensor in our program, no matter in prediction or training…. Read more »