Celong Liu

FSVideo Team, Qingyu Chen, Zhiyuan Fang et al.

We introduce FSVideo, a fast speed transformer-based image-to-video (I2V) diffusion framework. We build our framework on the following key components: 1.) a new video autoencoder with highly-compressed latent space ($64\times64\times4$ spatial-temporal downsampling ratio), achieving competitive reconstruction quality; 2.) a diffusion transformer (DIT) architecture with a new layer memory design to enhance inter-layer information flow and context reuse within DIT, and 3.) a multi-resolution generation strategy via a few-step DIT upsampler to increase video fidelity. Our final model, which contains a 14B DIT base model and a 14B DIT upsampler, achieves competitive performance against other popular open-source models, while being an order of magnitude faster. We discuss our model design as well as training strategies in this report.

Luchuan Song, Lele Chen, Celong Liu et al.

We propose TextToon, a method to generate a drivable toonified avatar. Given a short monocular video sequence and a written instruction about the avatar style, our model can generate a high-fidelity toonified avatar that can be driven in real-time by another video with arbitrary identities. Existing related works heavily rely on multi-view modeling to recover geometry via texture embeddings, presented in a static manner, leading to control limitations. The multi-view video input also makes it difficult to deploy these models in real-world applications. To address these issues, we adopt a conditional embedding Tri-plane to learn realistic and stylized facial representations in a Gaussian deformation field. Additionally, we expand the stylization capabilities of 3D Gaussian Splatting by introducing an adaptive pixel-translation neural network and leveraging patch-aware contrastive learning to achieve high-quality images. To push our work into consumer applications, we develop a real-time system that can operate at 48 FPS on a GPU machine and 15-18 FPS on a mobile machine. Extensive experiments demonstrate the efficacy of our approach in generating textual avatars over existing methods in terms of quality and real-time animation. Please refer to our project page for more details: https://songluchuan.github.io/TextToon/.

Lele Chen, Guofeng Cui, Celong Liu et al.

When people deliver a speech, they naturally move heads, and this rhythmic head motion conveys prosodic information. However, generating a lip-synced video while moving head naturally is challenging. While remarkably successful, existing works either generate still talkingface videos or rely on landmark/video frames as sparse/dense mapping guidance to generate head movements, which leads to unrealistic or uncontrollable video synthesis. To overcome the limitations, we propose a 3D-aware generative network along with a hybrid embedding module and a non-linear composition module. Through modeling the head motion and facial expressions1 explicitly, manipulating 3D animation carefully, and embedding reference images dynamically, our approach achieves controllable, photo-realistic, and temporally coherent talking-head videos with natural head movements. Thoughtful experiments on several standard benchmarks demonstrate that our method achieves significantly better results than the state-of-the-art methods in both quantitative and qualitative comparisons. The code is available on https://github.com/ lelechen63/Talking-head-Generation-with-Rhythmic-Head-Motion.

Zhong Li, Liangchen Song, Celong Liu et al.

In this paper, we present an efficient and robust deep learning solution for novel view synthesis of complex scenes. In our approach, a 3D scene is represented as a light field, i.e., a set of rays, each of which has a corresponding color when reaching the image plane. For efficient novel view rendering, we adopt a two-plane parameterization of the light field, where each ray is characterized by a 4D parameter. We then formulate the light field as a 4D function that maps 4D coordinates to corresponding color values. We train a deep fully connected network to optimize this implicit function and memorize the 3D scene. Then, the scene-specific model is used to synthesize novel views. Different from previous light field approaches which require dense view sampling to reliably render novel views, our method can render novel views by sampling rays and querying the color for each ray from the network directly, thus enabling high-quality light field rendering with a sparser set of training images. Per-ray depth can be optionally predicted by the network, thus enabling applications such as auto refocus. Our novel view synthesis results are comparable to the state-of-the-arts, and even superior in some challenging scenes with refraction and reflection. We achieve this while maintaining an interactive frame rate and a small memory footprint.

Celong Liu, Xin Li

We develop a new algorithm to perform facial reconstruction from a given skull. This technique has forensic application in helping the identification of skeletal remains when other information is unavailable. Unlike most existing strategies that directly reconstruct the face from the skull, we utilize a database of portrait photos to create many face candidates, then perform a superimposition to get a well matched face, and then revise it according to the superimposition. To support this pipeline, we build an effective autoencoder for image-based facial reconstruction, and a generative model for constrained face inpainting. Our experiments have demonstrated that the proposed pipeline is stable and accurate.