Search arXiv⌕ Search

arXiv subjects

Renjie Zou

Publications and source records attributed to Renjie Zou.

5 recordsLinked to original sources

Noise-Regularized Training for Learned Image Compression

Learned image compression (LIC) is trained almost exclusively on clean images: the only stochastic perturbation in the loop is an additive uniform proxy injected in the latent domain to keep the entropy model differentiable. Outside compression, by contrast, a long line of work---from Tikhonov regularization and contractive autoencoders through denoising score matching and modern diffusion models---has established that Gaussian noise at the input of a network with a clean target acts as a structured probe of the data manifold rather than as data augmentation. Building on this view, we introduce noise-regularized training for LIC: at every training step we add Gaussian noise to the encoder input while keeping the distortion target, the latent uniform proxy, the hyperprior, and the entropy coding pipeline unchanged. The recipe is architecture-agnostic and adds essentially no training or inference overhead. Across backbones and operating regimes it yields consistent BD-Rate gains: -2.60% on a reproduced ELIC and -0.85% / -1.05% / -2.38% on the Base/Medium/Large scales of a variable-rate DMCI codec, with DMCI-Large reaching -22.02% against VTM-22.0. A second-order expansion of the noisy RD loss accounts for these gains: because the noise enters upstream of the analysis transform, it converts the clean RD objective into a structured $O(σ^{2})$ regularizer comprising an information-weighted encoder-Jacobian penalty (rate side) and a Frobenius contractive penalty on the end-to-end reconstruction map (distortion side). A finite-difference Lipschitz analysis shows corresponding sensitivity reductions of ~12% for the encoder and ~8% for the end-to-end codec, while an exact-JVP test further shows preferential contraction of high-information directions. Code and trained models are available at https://github.com/Googolxx/NoiseRegu-LIC.

eess.IV↗

Extraction of All Partial-Wave Interaction Parameters via Femtoscopy: The Central-Force Case

We present a universal femtoscopic method to extract the source size and central-force interaction parameters for arbitrary partial waves from experimental data, with the interaction parameters defined in the effective-range expansion. The method is based on the Wronskian determinant and is applicable to both the Coulomb and non-Coulomb cases. This approach is numerically verified against simulated neutron-neutron and proton-proton correlation data obtained from the Reid potential.

nucl-th↗

Extended Variable Phase Method for Spin-1/2 Correlation Functions

We have developed a systematic approach to calculate the correlation function for spin-1/2 particles, incorporating both central and noncentral components of the interparticle interaction. This is achieved by extending the variable phase method to accommodate noncentral potentials and numerically solving the Schrödinger equation. Within this framework, the partial-wave contributions to the nucleon-nucleon correlation functions adopting the Reid soft-core potential are evaluated. The resulting correlation functions are then compared for Gaussian sources of different sizes.

nucl-th↗

Imaging Freeze-out Sources and Extracting Strong Interaction Parameters in Relativistic Heavy-Ion Collisions

By combining femtoscopic interferometry with an optical deblurring algorithm, we present a novel method to image the source in heavy-ion collisions (HICs) while simultaneously determining the interaction strength between particle pairs. The spatial distribution of the emission source has been reconstructed for protons ($p$) and antiprotons ($\bar{p}$) from the respective $pp$ and $\bar{p}\bar{p}$ correlation functions in Au+Au collisions at $\sqrt{S_{\rm NN}}=200$ GeV. Within experimental uncertainties, protons and antiprotons share the same freeze-out distribution showing higher density in the center compared to the widely assumed Gaussian shape. The results evidence the matter-antimatter symmetry in coordinate space at the moment of freeze-out before the nucleons are fully randomized in the collisions.

nucl-th↗

The Devil Is in the Details: Window-based Attention for Image Compression

Learned image compression methods have exhibited superior rate-distortion performance than classical image compression standards. Most existing learned image compression models are based on Convolutional Neural Networks (CNNs). Despite great contributions, a main drawback of CNN based model is that its structure is not designed for capturing local redundancy, especially the non-repetitive textures, which severely affects the reconstruction quality. Therefore, how to make full use of both global structure and local texture becomes the core problem for learning-based image compression. Inspired by recent progresses of Vision Transformer (ViT) and Swin Transformer, we found that combining the local-aware attention mechanism with the global-related feature learning could meet the expectation in image compression. In this paper, we first extensively study the effects of multiple kinds of attention mechanisms for local features learning, then introduce a more straightforward yet effective window-based local attention block. The proposed window-based attention is very flexible which could work as a plug-and-play component to enhance CNN and Transformer models. Moreover, we propose a novel Symmetrical TransFormer (STF) framework with absolute transformer blocks in the down-sampling encoder and up-sampling decoder. Extensive experimental evaluations have shown that the proposed method is effective and outperforms the state-of-the-art methods. The code is publicly available at https://github.com/Googolxx/STF.

eess.IV↗