object-detection

Dear all,

I hope to find you well. The current codebase is totally closed and it is impossible to retrieve all the metrics computed. Allow all metrics to be returned inside dataset.evaluate. For instance, currently, we cannot get the map per class that is printed inside [eval_map](https://github.com/open-mmlab/mmdetection/blob/414c62cb12d1b0ebf6151a697a32b84adc269ca4/mmdet/core/evaluati

Enhancement

A discussion in #614 revealed a good place for improvement - we should ensure that input image is continuous upon start of the augmentation pipeline. This could be implemented by adding image = np.ascontiguousarray(image) to image and mask targets.

A proposed place to add this call - somewhere at the beginning of A.Compose.__call__.

There are many links in Kinetics that have expired. As as result, everyone might not be using the same Kinetics dataset. As a reference, the statistics of the Kinetics dataset used in PySlowFast can be found here, https://github.com/facebookresearch/video-nonlocal-net/blob/master/DATASET.md. However, I cannot seem to find similar information for gluoncv. Will you guys be sharing the statistics and

我发现很多人在使用voc格式的数据集时，和我遇到了同样的问题，训练时AP一直为0，
今早，仔细检查后，我也找到了真正的原因，主要是数据加载的地方出现了问题，还是我们自己太不仔细了

解决流程思路： 解决YOLOX训练时AP为0

When running TabularPredictor.fit(), I encounter a BrokenPipeError for some reason.
What is causing this?
Could it be due to OOM error?

Fitting model: XGBoost ...
-34.1179 = Validation root_mean_squared_error score
10.58s = Training runtime
0.03s = Validation runtime
Fitting model: NeuralNetMXNet ...
-34.2849 = Validation root_mean_squared_error score
43.63s =