Pan Wang
commited on
Commit
·
19284a5
1
Parent(s):
088408d
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,5 +1,7 @@
|
|
| 1 |
# DLF: Disentangled-Language-Focused Multimodal Sentiment Analysis, AAAI 2025.
|
| 2 |
|
|
|
|
|
|
|
| 3 |
## Main Contributions
|
| 4 |
|
| 5 |
Our main contributions can be summarized as follows:
|
|
@@ -11,7 +13,7 @@ Our main contributions can be summarized as follows:
|
|
| 11 |
|
| 12 |
## The Framework
|
| 13 |

|
| 14 |
-
The framework of DLF. Please refer to [Paper Link](arxiv) for details.
|
| 15 |
|
| 16 |
|
| 17 |
## Usage
|
|
@@ -41,13 +43,13 @@ git clone https://github.com/pwang322/DLF.git
|
|
| 41 |
- Install the necessary packages.
|
| 42 |
```
|
| 43 |
cd DLF
|
| 44 |
-
pip install requirements.txt
|
| 45 |
```
|
| 46 |
|
| 47 |
### Datasets
|
| 48 |
Data files (containing processed MOSI, MOSEI datasets) can be downloaded from [here](https://drive.google.com/drive/folders/1BBadVSptOe4h8TWchkhWZRLJw8YG_aEi?usp=sharing).
|
| 49 |
You can first build and then put the downloaded datasets into `./dataset` directory and revise the path in `./config/config.json`. For example, if the processed the MOSI dataset is located in `./dataset/MOSI/aligned_50.pkl`. Please make sure "dataset_root_dir": "./dataset" and "featurePath": "MOSI/aligned_50.pkl".
|
| 50 |
-
Please note that the meta information and the raw data are not available due to privacy of
|
| 51 |
|
| 52 |
### Run the Codes
|
| 53 |
- Training
|
|
@@ -74,7 +76,7 @@ If you find the code and our idea helpful in your research or work, please cite
|
|
| 74 |
@article{wang2025dlf,
|
| 75 |
title={DLF: Disentangled-Language-Focused Multimodal Sentiment Analysis},
|
| 76 |
author={Wang, Pan and Zhou, Qiang and Wu, Yawen and Chen, Tianlong and Hu, Jingtong},
|
| 77 |
-
journal={arXiv preprint arXiv:2412},
|
| 78 |
year={2024}
|
| 79 |
}
|
| 80 |
```
|
|
|
|
| 1 |
# DLF: Disentangled-Language-Focused Multimodal Sentiment Analysis, AAAI 2025.
|
| 2 |
|
| 3 |
+
### [Arxiv Paper](https://arxiv.org/abs/2412.12225)
|
| 4 |
+
|
| 5 |
## Main Contributions
|
| 6 |
|
| 7 |
Our main contributions can be summarized as follows:
|
|
|
|
| 13 |
|
| 14 |
## The Framework
|
| 15 |

|
| 16 |
+
The framework of DLF. Please refer to [Paper Link](https://arxiv.org/abs/2412.12225) for details.
|
| 17 |
|
| 18 |
|
| 19 |
## Usage
|
|
|
|
| 43 |
- Install the necessary packages.
|
| 44 |
```
|
| 45 |
cd DLF
|
| 46 |
+
pip install -r requirements.txt
|
| 47 |
```
|
| 48 |
|
| 49 |
### Datasets
|
| 50 |
Data files (containing processed MOSI, MOSEI datasets) can be downloaded from [here](https://drive.google.com/drive/folders/1BBadVSptOe4h8TWchkhWZRLJw8YG_aEi?usp=sharing).
|
| 51 |
You can first build and then put the downloaded datasets into `./dataset` directory and revise the path in `./config/config.json`. For example, if the processed the MOSI dataset is located in `./dataset/MOSI/aligned_50.pkl`. Please make sure "dataset_root_dir": "./dataset" and "featurePath": "MOSI/aligned_50.pkl".
|
| 52 |
+
Please note that the meta information and the raw data are not available due to the privacy of YouTube content creators. For more details, please follow the [official website](https://github.com/ecfm/CMU-MultimodalSDK) of these datasets.
|
| 53 |
|
| 54 |
### Run the Codes
|
| 55 |
- Training
|
|
|
|
| 76 |
@article{wang2025dlf,
|
| 77 |
title={DLF: Disentangled-Language-Focused Multimodal Sentiment Analysis},
|
| 78 |
author={Wang, Pan and Zhou, Qiang and Wu, Yawen and Chen, Tianlong and Hu, Jingtong},
|
| 79 |
+
journal={arXiv preprint arXiv:2412.12225},
|
| 80 |
year={2024}
|
| 81 |
}
|
| 82 |
```
|