13 papers
SoLAR: Error-Resilient Streamable Long-Horizon Free-Viewpoint Video Reconstruction with Anchor Activation and Latent Recalibration
Haotian Zhang, Xu Mo, Yixin Yu +7
Free-Viewpoint Video (FVV) has emerged as a cornerstone of next-generation immersive media systems and attracted widespread attention. Previous methods primarily focus on short vid…
Training-Free Image Editing with Visual Context Integration and Concept Alignment
Rui Song, Guo-Hua Wang, Qing-Guo Chen +6
In image editing, it is essential to incorporate a context image to convey the user's precise requirements, such as subject appearance or image style. Existing training-based visua…
RAWIC: Bit-Depth Adaptive Lossless Raw Image Compression
Chunhang Zheng, Tongda Xu, Mingli Xie +2
Raw images preserve linear sensor measurements and high bit-depth information crucial for advanced vision tasks and photography applications, yet their storage remains challenging…
Parallax to Align Them All: An OmniParallax Attention Mechanism for Distributed Multi-View Image Compression
Haotian Zhang, Feiyue Long, Yixin Yu +7
Multi-view image compression (MIC) aims to achieve high compression efficiency by exploiting inter-image correlations, playing a crucial role in 3D applications. As a subfield of M…
Making Reconstruction FID Predictive of Diffusion Generation FID
Tongda Xu, Mingwei He, Shady Abu-Hussein +6
It is well known that the reconstruction FID (rFID) of a VAE is poorly correlated with the generation FID (gFID) of a latent diffusion model. We propose interpolated FID (iFID), a…
Benchmarking and Enhancing VLM for Compressed Image Understanding
Zifu Zhang, Tongda Xu, Siqi Li +4
With the rapid development of Vision-Language Models (VLMs) and the growing demand for their applications, efficient compression of the image inputs has become increasingly importa…
Training-Free Vector Quantization via Gaussian VAEs
Tongda Xu, Wendi Zheng, Jiajun He +4
Vector-quantized variational autoencoders (VQ-VAEs) are discrete autoencoders that compress images into discrete tokens. However, they are difficult to train due to discretization.…
V-Shuffle: Zero-Shot Style Transfer via Value Shuffle
Haojun Tang, Qiwei Lin, Tongda Xu +2
Attention injection-based style transfer has achieved remarkable progress in recent years. However, existing methods often suffer from content leakage, where the undesired semantic…
Versatile Recompression-Aware Perceptual Image Super-Resolution
Mingwei He, Tongda Xu, Xingtong Ge +3
Perceptual image super-resolution (SR) methods restore degraded images and produce sharp outputs. In practice, those outputs are usually recompressed for storage and transmission.…
PICD: Versatile Perceptual Image Compression with Diffusion Rendering
Tongda Xu, Jiahao Li, Bin Li +3
Recently, perceptual image compression has achieved significant advancements, delivering high visual quality at low bitrates for natural images. However, for screen content, existi…
CoopDETR: A Unified Cooperative Perception Framework for 3D Detection via Object Query
Zhe Wang, Shaocong Xu, Xucai Zhuang +5
Cooperative perception enhances the individual perception capabilities of autonomous vehicles (AVs) by providing a comprehensive view of the environment. However, balancing percept…
Noise Dimension of GAN: An Image Compression Perspective
Ziran Zhu, Tongda Xu, Ling Li +1
Generative adversial network (GAN) is a type of generative model that maps a high-dimensional noise to samples in target distribution. However, the dimension of noise required in G…
Consistency Model is an Effective Posterior Sample Approximation for Diffusion Inverse Solvers
Tongda Xu, Ziran Zhu, Jian Li +8
Diffusion Inverse Solvers (DIS) are designed to sample from the conditional distribution , with a predefined diffusion model , an operator , and a m…