-
- https://arxiv.org/abs/2107.13731
- IJCAI 2021
スマートデバイスのアクセシビリティを向上させ、その利用を容易にするためには、ユーザーインターフェース(UI)を理解し、ユーザーがタスクを完了するのを支援するモデルを構築することが重要です。
しかし、画像、テキスト、構造的なメタデータを含むマルチモーダルなUIの特徴を効果的に活用する方法や、…
e4exp updated
3 years ago
-
Thanks for your work on ReID! I have a few questions:
1、I would like to ask if this combination of global and local vit as backbone has been tried at ClusterContrtast?
2、When calculating…
-
Hi
What do you think about adding loss functions in contrastive learning such as N-pair loss (https://papers.nips.cc/paper/2016/hash/6b180037abbebea991d8b1232f8a8ca9-Abstract.html, https://lilianweng…
-
![image](https://github.com/IDEA-Research/GroundingDINO/assets/14495189/e070e6f0-3156-432a-8902-6ba37dbd1bcb)
作者您好,下面所有的描述都是以上图中红框的设置进行实验的。
我在复现groundingding的训练代码,用作者开源的swin-t在O365,GoldG,Cap4M上预训练模型…
-
Hi! may you provide the training log of the experiment? I want to see the convergence curve of contrastive loss. Because the implemented contrastive loss by myself borrows largely from your code but d…
-
-
Hi mate,
Thanks for publicizing this exellent work. I skimmed through the code and found that you took norm twice while calculating the patch level contrastive loss here https://github.com/cyclomon/O…
-
### System Info
- `transformers` version: 4.37.0.dev0
- Platform: Linux-5.15.0-88-generic-x86_64-with-glibc2.31
- Python version: 3.11.5
- Huggingface_hub version: 0.20.1
- Safetensors version: 0…
-
Hi! What is the difference between Contrast_Loss_ds and Contrast_Loss? Plus may you provide the training log of the experiment? I want to see the convergence curve of contrastive loss. Thanks a lot!
-
**System information**.
TensorFlow version (you are using): 2.12
Are you willing to contribute it (Yes/No) : Yes
**Describe the feature and the current behavior/state**.
Pair loss is used in …