-
The discriminative step is now limited to the capacity of one GPU
We must be able to perform the training of all single models on several GPUs.
-
In this code of eigen descriotor, order of eigen value is e1
-
Line #54 in Discriminative-Sounding-Objects-Localization/music-exp/data/cut_audios.py
-
Paper link: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6624009
-
Institute: Sydney
URL: https://arxiv.org/pdf/1805.03508.pdf
Keyword: Visual Grounding
Interest: 3
> VQA 기반의 Dialog를 향해 나아가기 위해서, object detector를 대체하게 될 방향이므로 반드시 팔로웝해야함
-
Want to know about base_cam.py , is this implementation of the earlier version of grad_CAM , I mean the normal Class Activation Maps based on paper "Learning Deep Features for Discriminative Localizat…
-
I am working on the 3DShapeNet now, however, I have no idea on the pretrained model, discriminative_40_class.mat and generative_40_class.mat.
Anyone can tell me where to find the pretrained models?…
-
![image](https://github.com/YuanGongND/ssast/assets/44318771/f8e85f24-d1a1-418f-b88a-e93c507d78c3)
In this piece of code, it seems that the loss is calculated at the granularity of samples.
-
#ACMMM2017
Keywords: Fine-grain, Localization
Institute: Peking, Beijing
URL: https://arxiv.org/pdf/1709.08295.pdf
Interest: 3
-
Are you using the [2014 DSST](http://www.cvl.isy.liu.se/en/research/objrec/visualtracking/scalvistrack/ScaleTracking_BMVC14.pdf) or the [2016 fDSST](https://arxiv.org/pdf/1609.06141.pdf) version of Di…