https://github.com/yfzhang114/llava-align

This is the official repo for Debiasing Large Visual Language Models, including a Post-Hoc debias method and Visual Debias Decoding strategy.
https://github.com/yfzhang114/llava-align

debiasing hallucination large-vision-language-models

Last synced: 6 months ago
JSON representation

This is the official repo for Debiasing Large Visual Language Models, including a Post-Hoc debias method and Visual Debias Decoding strategy.

• Host: GitHub
• URL: https://github.com/yfzhang114/llava-align
• Owner: yfzhang114
• License: apache-2.0
• Created: 2024-01-23T03:30:18.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-02-22T02:07:26.000Z (8 months ago)
• Last Synced: 2025-04-10T01:13:24.965Z (6 months ago)
• Topics: debiasing, hallucination, large-vision-language-models
• Language: Python
• Homepage:
• Size: 64.9 MB
• Stars: 77
• Watchers: 1
• Forks: 2
• Open Issues: 3
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          


# Debiasing Large Visual Language Models / Debiasing Multimodal Large Language Models

This is the official repo for Debiasing Large Visual Language Models, including a Post-Hoc debias method and Visual Debias Decoding strategy. These strategies not only prove beneficial in minimizing hallucinations but also contribute to the generation of more helpful and precise illustrations

## 🔥 Update

* [2024-03-08]: ⭐️ Paper online. Check out [Debiasing Multimodal Large Language Models](https://arxiv.org/abs/2403.05262) for details.

* [2024-03-11]: 🚀🚀 Codes released.

## 🎯 Overview

![LLaVA-v1.5-7B generate confident answer with meaningless images. "None" indicates the absence of an input image, while "Noise" signifies the presence of Gaussian noise matching the image dimensions. "Zeros, Ones" indicates a scenario where a tensor with all zero/one values.](figs/model_bias.png)

- Our investigation reveals a noteworthy bias in the generated content, where the output is primarily influenced by the underlying Language Models (LLMs) rather than the input image.

![Large Visual Language Models Debiasing](figs/bias_model.png)

- We introduce Post-Hoc debias, where a "calibration" step is implemented for the model's output probabilities using an affine transformation.

- We introduce Visual Debias Decoding (VDD), **a simple and training-free** method that contrasts output distributions derived from original and image-free visual inputs.

- These strategies not only prove beneficial in minimizing hallucinations but also contribute to the generation of more helpful and precise illustrations.

## 🕹️ Usage

### Environment Setup

```bash

conda create -yn vdd python=3.9

conda activate vdd

cd LLaVA-Align

pip install -r requirements.txt

```

### Re-implementation of Our Results

For all experiments presented in our paper, refer to the `experiments/scripts` directory for detailed commands and scripts. Below, we provide simple implementation examples and guidance.

### How to Use Post-Hoc Debiasing in LVLMs

To implement Post-Hoc Debiasing in LVLMs, follow these steps:

1. **Obtain Output Distributions:**

   - Generate output distributions with naive image and meaningless vision information, such as pure text input (None) or by replacing vision tokens with ``.

   - For the POPE benchmark, use the following code to calculate the top-k tokens with their probabilities for each input:

     ```bash

     ./eval/calibrate/llava_calibrate.py

     ```

2. **Initialize Debiasing Weight:**

   - With the obtained naive classification distribution and debiased classification distribution, initialize the debiasing weight $W$ and bias $b$.

   - Adjust the output distribution using affine transformation.

3. **Concrete Example - POPE Binary Classification:**

   - For the POPE binary classification setting, run the following code to see a concrete example:

     ```bash

     ./eval/eval_pope_calibrate.py

     ```

Feel free to modify the parameters or refer to the code for additional details on the implementation of Post-Hoc Debiasing in LVLMs.

### How to Use VDD in LVLMs

To help you get started quickly, here's an example using LLaVA on how to replace the conventional sampling method with the VDD/VCD method during generation:

1. Add the following at the beginning of the start-up script:

```python

from vcd_utils.vcd_sample import evolve_vcd_sampling

evolve_vcd_sampling()

```

The `evolve_vcd_sampling` function replaces the sampling function in the transformers library. The modified sampling function includes an option for visual contrastive decoding, while keeping the rest unchanged.

2. Slightly modify `llava_llama.py`:

   a. Add contrastive decoding parameters in the `LlavaLlamaForCausalLM` class's `forward` function to avoid exceptions in `model.generate`.

   

   b. Add the `prepare_inputs_for_generation_cd` function.

3. set the hyperparameter in the `generate` function:

```python

output_ids = model.generate(

    input_ids,

    images=image_tensor.unsqueeze(0).half().cuda(),

    use_dd=args.use_dd,

    use_dd_unk=args.use_dd_unk,

    cd_alpha = args.cd_alpha,

    cd_beta = args.cd_beta,

    do_sample=True)

```

```

--use_dd: use pure text input for debias decoding

--use_dd_unk: replace the image tokens with  token, the output logits will be used for debias decoding

--use_dd --use_dd_unk: use both of them

```

### How to Test the Effects of Decoding Configurations on LVLMs

To assess the impact of decoding configurations on LVLMs, follow these steps:

1. **Implement For-Loops:**

   - Implement for-loops on the temperature, top-p, and top-k configurations in the decoding process.

   - Collect the results obtained for each configuration.

2. **Evaluate Results:**

   - Evaluate the collected results using a similar approach as described in the preceding sections.

3. **Concrete Example:**

   - For a concrete example for POPE or LLaVA-Bench, run the following code to see how the implementation works:

     ```bash

     ./eval/sampling/llava_sampling.py

     ```

## 📌 Examples

![Case2](figs/sample2.png)

*Figure 13. Qualitative examples showcasing the impact of VDD on LLaVA-v1.5-13B. VDD is demonstrated to be less hallucinated.*

![Case3](figs/sample3.png)

*Figure 14. Another Qualitative examples showcasing the impact of VDD on LLaVA-v1.5-13B. VDD is demonstrated to be more helpful and precies.*

## 📑 Citation

If you find our project useful, we hope you can star our repo and cite our paper as follows:

```

@article{zhang2024debiasing,

  title={Debiasing multimodal large language models},

  author={Zhang, Yi-Fan and Yu, Weichen and Wen, Qingsong and Wang, Xue and Zhang, Zhang and Wang, Liang and Jin, Rong and Tan, Tieniu},

  journal={arXiv preprint arXiv:2403.05262},

  year={2024}

}

```

## 📝 Related Projects

- [Contrastive Decoding](https://github.com/XiangLi1999/ContrastiveDecoding): Open-ended Text Generation as Optimization

- [Qwen-VL](https://github.com/QwenLM/Qwen-VL): A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond

- [LLaVA 1.5](https://github.com/haotian-liu/LLaVA): Improved Baselines with Visual Instruction Tuning

- [VCD](https://github.com/DAMO-NLP-SG/VCD/): Visual Contrastive Decoding