Update README.md

README.md

@@ -10,36 +10,31 @@ inference: false
 
 # Trajectory Consistency Distillation
 
-[![Arxiv](https://img.shields.io/badge/arXiv-2402.xxxxx-b31b1b)]()
-[![Project page](https://img.shields.io/badge/Web-Project%20Page-green)](https://mhh0318.github.io/tcd/)
-[![Github](https://img.shields.io/badge/Github-Repo-yellow?logo=github)](https://github.com/jabir-zheng/TCD)
-[![Hugging Face Space](https://img.shields.io/badge/%F0%9F%A4%97HuggingFace-Space-blue)]()
+Official Model Repo of the paper: [Trajectory Consistency Distillation](https://arxiv.org/abs/2402.19159).
+For more information, please check the [GitHub Repo](https://github.com/jabir-zheng/TCD) and [Project Page](https://mhh0318.github.io/tcd/).
 
-Official Repository of the paper: [Trajectory Consistency Distillation]()
+Also welcome to try the demo host on [🤗 Space(https://huggingface.co/spaces/h1t/TCD)].
 
 ![](./assets/teaser_fig.png)
 
-## News
-- (🔥New) 2024/2/28 We provided a demo of TCD on 🤗 Hugging Face Space. Try it out [here]().
-- (🔥New) 2024/2/28 We released our model [TCD-SDXL-Lora]() in 🤗 Hugging Face.
-- (🔥New) 2024/2/28 Please refer to the [Usage](#usage-anchor) for more information with Diffusers Pipeline.
-
 ## Introduction
 
 TCD, inspired by [Consistency Models](https://arxiv.org/abs/2303.01469), is a novel distillation technology that enables the distillation of knowledge from pre-trained diffusion models into a few-step sampler. In this repository, we release the inference code and our model named TCD-SDXL, which is distilled from [SDXL Base 1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0). We provide the LoRA checkpoint in this [repository]().
 
+![](./assets/teaser.jpeg)
+
 ✨TCD has following advantages:
 
-- `High-Quality with Few-Step`: TCD significantly surpasses the previous state-of-the-art few-step text-to-image model [LCM](https://github.com/luosiallen/latent-consistency-model/tree/main) in terms of image quality. Notably, LCM experiences a notable decline in quality at high NFEs. In contrast, _**TCD maintains superior generative quality at high NFEs, even exceeding the performance of DPM-Solver++(2S) with origin SDXL**_. 
-![](./assets/teaser.jpeg)
-<!-- We observed that the images generated with 8 steps by TCD-SDXL are already highly impressive, even outperforming the original SDXL 50-steps generation results. -->
+- `Flexible NFEs`: For TCD, the NFEs can be varied at will (compared with Turbo), without adversely affecting the quality of the results (compared with LCMs), where LCM experiences a notable decline in quality at high NFEs. 
+- `Better than Teacher`: TCD maintains superior generative quality at high NFEs, even exceeding the performance of DPM-Solver++(2S) with origin SDXL. It is worth noting that there is no additional discriminator or LPIPS supervision included during training.
+- `Freely Change the Detailing`: During inference, the level of detail in the image can be simply modified by adjusing one hyper-parameter gamma. This option does not require the introduction of any additional parameters.
 - `Versatility`: Integrated with LoRA technology, TCD can be directly applied to various models (including the custom Community Models, styled LoRA, ControlNet, IP-Adapter) that share the same backbone, as demonstrated in the [Usage](#usage-anchor).
 ![](./assets/versatility.png)
 - `Avoiding Mode Collapse`: TCD achieves few-step generation without the need for adversarial training, thus circumventing mode collapse caused by the GAN objective. 
 In contrast to the concurrent work [SDXL-Lightning](https://huggingface.co/ByteDance/SDXL-Lightning), which relies on Adversarial Diffusion Distillation, TCD can synthesize results that are more realistic and slightly more diverse, without the presence of "Janus" artifacts.
 ![](./assets/compare_sdxl_lightning.png)
 
-For more information, please refer to our paper [Trajectory Consistency Distillation]().
+For more information, please refer to our paper [Trajectory Consistency Distillation](https://arxiv.org/abs/2402.19159).
 
 <a id="usage-anchor"></a>
 
@@ -318,6 +313,14 @@ grid_image = make_image_grid([ref_image, image], rows=1, cols=2)
 
 ## Citation
 ```bibtex
+@misc{zheng2024trajectory,
+      title={Trajectory Consistency Distillation}, 
+      author={Jianbin Zheng and Minghui Hu and Zhongyi Fan and Chaoyue Wang and Changxing Ding and Dacheng Tao and Tat-Jen Cham},
+      year={2024},
+      eprint={2402.19159},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV}
+}
 ```
 
 ## Acknowledgments