Jian Cheng

when i train a batch(batchsize 100) of flickr8k, it takes around 0.8 seconds,

4

when i train a batch(batchsize 100) of flickr8k, it takes around 0.8 second. However when train a batch(still 100) of flickr30k, it takes around 7 seconds. Why?

注意到作者提供的yolov3.prototxt里面，layer84-route和layer96-route作为concat层，但是其本身只接受一个输入，请问这样做的用意是什么呀？

谢谢您的分享！请问有没有caffe版本的实现啊？

failed after loading images and superpixel segmentation

5

After 'loading images' and 'superpixel segmentation', the process will be killed, and the error is 'can't get the stat of '_tmp/smaps_sc/*: no that file or directory''.. Did i miss something...

Why model output pred_score is all zeros?

4

when inferencing set predictor model, why the output pred_score is all zerors? I notice that pred_probs is initialized with 'np.zeros((batch_size, 6))' in function: goals_2D_eval() if 'set_predict', and as a output...

Hi, I have a question for nuscenes prediction, does LAformer only predict ego vehicle's trajectory?