-
It is known that for a specific sample(query), its negative samples are all from dictionary(queue). I am wondering if adding other keys in the same batch as negative samples will make sense or not?
…
-
Anyone has code for custom training on own dataset???
-
Hi, author!
Thanks for your sharing! You do an impressive work!
I have two question.
The first is what will happen if CLIP image representation is used to replace SSL representation in the first tw…
-
Does anyone know how to generate the visual results in Figure 6? I see that they extract SSL representations from image samples, and the authors don't seem to describe how they combine these features …
-
## 背景
PyTorch 最近在 2.1.0 版本发布了 `torch.export` 功能，提供了以 PyTorch 2.x 的 dynamo 为核心的模型导出方法。其功能上与我们 Paddle 动转静模型导出 `paddle.jit.save` 相类似，因此我们想要建设一个 repo 能够自动对比 `torch.export` 和 `paddle.jit.save` 的导出成功与否，[…
-
- https://arxiv.org/abs/2105.04553
- 2021
コンピュータビジョンの分野では、CNNからTransformerへとモデルチェンジが進んでいます。
本研究では、Vision Transformersをバックボーン・アーキテクチャとする、MoBYと呼ばれる自己教師付き学習アプローチを発表する。
このアプローチには基本的に新しい発明はなく、MoCo v2…
e4exp updated
3 years ago
-
### 🐛 Describe the bug
Note: i'm not duplicating issues across backends. A failure with inductor will also occur with cudagraphs and dynamic, and a cudagraphs failure will occur with cudagraphs dyn…
-
### Prerequisite
- [X] I have searched [the existing and past issues](https://github.com/open-mmlab/mmyolo/issues) but cannot get the expected help.
- [X] I have read the [FAQ documentation](https://…
-
1. Use metadata.py to query these metadata
2. In CI, check each model has these metadata
-
# Contrastive Learning
## InstDisc, CVPR 2018
Unsupervised Feature Learning via Non-Parametric Instance Discrimination, CVPR 2018. [[paper](https://openaccess.thecvf.com/content_cvpr_2018/html/W…