Heming Sun

Jinming Liu, Heming Sun, Jiro Katto

Learned image compression (LIC) methods have exhibited promising progress and superior rate-distortion performance compared with classical image compression standards. Most existing LIC methods are Convolutional Neural Networks-based (CNN-based) or Transformer-based, which have different advantages. Exploiting both advantages is a point worth exploring, which has two challenges: 1) how to effectively fuse the two methods? 2) how to achieve higher performance with a suitable complexity? In this paper, we propose an efficient parallel Transformer-CNN Mixture (TCM) block with a controllable complexity to incorporate the local modeling ability of CNN and the non-local modeling ability of transformers to improve the overall architecture of image compression models. Besides, inspired by the recent progress of entropy estimation models and attention modules, we propose a channel-wise entropy model with parameter-efficient swin-transformer-based attention (SWAtten) modules by using channel squeezing. Experimental results demonstrate our proposed method achieves state-of-the-art rate-distortion performances on three different resolution datasets (i.e., Kodak, Tecnick, CLIC Professional Validation) compared to existing LIC methods. The code is at https://github.com/jmliu206/LIC_TCM.

Ao Luo, Linxin Song, Keisuke Nonaka et al.

In recent years, the task of learned point cloud compression has gained prominence. An important type of point cloud, the spinning LiDAR point cloud, is generated by spinning LiDAR on vehicles. This process results in numerous circular shapes and azimuthal angle invariance features within the point clouds. However, these two features have been largely overlooked by previous methodologies. In this paper, we introduce a model-agnostic method called Spherical-Coordinate-based learned Point cloud compression (SCP), designed to leverage the aforementioned features fully. Additionally, we propose a multi-level Octree for SCP to mitigate the reconstruction error for distant areas within the Spherical-coordinate-based Octree. SCP exhibits excellent universality, making it applicable to various learned point cloud compression techniques. Experimental results demonstrate that SCP surpasses previous state-of-the-art methods by up to 29.14% in point-to-point PSNR BD-Rate.

Fangzheng Lin, Heming Sun, Jinming Liu et al.

Recent state-of-the-art Learned Image Compression methods feature spatial context models, achieving great rate-distortion improvements over hyperprior methods. However, the autoregressive context model requires serial decoding, limiting runtime performance. The Checkerboard context model allows parallel decoding at a cost of reduced RD performance. We present a series of multistage spatial context models allowing both fast decoding and better RD performance. We split the latent space into square patches and decode serially within each patch while different patches are decoded in parallel. The proposed method features a comparable decoding speed to Checkerboard while reaching the RD performance of Autoregressive and even also outperforming Autoregressive. Inside each patch, the decoding order must be carefully decided as a bad order negatively impacts performance; therefore, we also propose a decoding order optimization algorithm.

Jinming Liu, Heming Sun, Jiro Katto

Most machine vision tasks (e.g., semantic segmentation) are based on images encoded and decoded by image compression algorithms (e.g., JPEG). However, these decoded images in the pixel domain introduce distortion, and they are optimized for human perception, making the performance of machine vision tasks suboptimal. In this paper, we propose a method based on the compressed domain to improve segmentation tasks. i) A dynamic and a static channel selection method are proposed to reduce the redundancy of compressed representations that are obtained by encoding. ii) Two different transform modules are explored and analyzed to help the compressed representation be transformed as the features in the segmentation network. The experimental results show that we can save up to 15.8\% bitrates compared with a state-of-the-art compressed domain-based work while saving up to about 83.6\% bitrates and 44.8\% inference time compared with the pixel domain-based method.

Fangzheng Lin, Heming Sun, Jiro Katto

Learned image compression allows achieving state-of-the-art accuracy and compression ratios, but their relatively slow runtime performance limits their usage. While previous attempts on optimizing learned image codecs focused more on the neural model and entropy coding, we present an alternative method to improving the runtime performance of various learned image compression models. We introduce multi-threaded pipelining and an optimized memory model to enable GPU and CPU workloads asynchronous execution, fully taking advantage of computational resources. Our architecture alone already produces excellent performance without any change to the neural model itself. We also demonstrate that combining our architecture with previous tweaks to the neural models can further improve runtime performance. We show that our implementations excel in throughput and latency compared to the baseline and demonstrate the performance of our implementations by creating a real-time video streaming encoder-decoder sample application, with the encoder running on an embedded device.

Tiesunlong Shen, Rui Mao, Jin Wang et al.

Aligning Large Language Models (LLMs) with human preferences is critical, yet traditional fine-tuning methods are computationally expensive and inflexible. While test-time alignment offers a promising alternative, existing approaches often rely on distorted trajectory-level signals or inefficient sampling, fundamentally capping performance and failing to preserve the generative diversity of the base model. This paper introduces LLMdoctor, a novel framework for efficient test-time alignment that operates via a patient-doctor paradigm. It integrates token-level reward acquisition with token-level flow-guided preference optimization (TFPO) to steer a large, frozen patient LLM with a smaller, specialized doctor model. Unlike conventional methods that rely on trajectory-level rewards, LLMdoctor first extracts fine-grained, token-level preference signals from the patient model's behavioral variations. These signals then guide the training of the doctor model via TFPO, which establishes flow consistency across all subtrajectories, enabling precise token-by-token alignment while inherently preserving generation diversity. Extensive experiments demonstrate that LLMdoctor significantly outperforms existing test-time alignment methods and even surpasses the performance of full fine-tuning approaches like DPO.

Jinming Liu, Yuntao Wei, Junyan Lin et al.

We present a new image compression paradigm to achieve ``intelligently coding for machine'' by cleverly leveraging the common sense of Large Multimodal Models (LMMs). We are motivated by the evidence that large language/multimodal models are powerful general-purpose semantics predictors for understanding the real world. Different from traditional image compression typically optimized for human eyes, the image coding for machines (ICM) framework we focus on requires the compressed bitstream to more comply with different downstream intelligent analysis tasks. To this end, we employ LMM to \textcolor{red}{tell codec what to compress}: 1) first utilize the powerful semantic understanding capability of LMMs w.r.t object grounding, identification, and importance ranking via prompts, to disentangle image content before compression, 2) and then based on these semantic priors we accordingly encode and transmit objects of the image in order with a structured bitstream. In this way, diverse vision benchmarks including image classification, object detection, instance segmentation, etc., can be well supported with such a semantically structured bitstream. We dub our method ``\textit{SDComp}'' for ``\textit{S}emantically \textit{D}isentangled \textit{Comp}ression'', and compare it with state-of-the-art codecs on a wide variety of different vision tasks. SDComp codec leads to more flexible reconstruction results, promised decoded visual quality, and a more generic/satisfactory intelligent task-supporting ability.

Shota Hirose, Shiori Maki, Naoki Wada et al.

Spectral Normalization is one of the best methods for stabilizing the training of Generative Adversarial Network. Spectral Normalization limits the gradient of discriminator between the distribution between real data and fake data. However, even with this normalization, GAN's training sometimes fails. In this paper, we reveal that more severe restriction is sometimes needed depending on the training dataset, then we propose a novel stabilizer which offers an adaptive normalization method, called ABCAS. Our method decides discriminator's Lipschitz constant adaptively, by checking the distance of distributions of real and fake data. Our method improves the stability of the training of Generative Adversarial Network and achieved better Fréchet Inception Distance score of generated images. We also investigated suitable spectral norm for three datasets. We show the result as an ablation study.

Ran Wang, Jinming Liu, Heming Sun et al.

Lossless image compression is an essential research field in image compression. Recently, learning-based image compression methods achieved impressive performance compared with traditional lossless methods, such as WebP, JPEG2000, and FLIF. However, there are still many impressive lossy compression methods that can be applied to lossless compression. Therefore, in this paper, we explore the methods widely used in lossy compression and apply them to lossless compression. Inspired by the impressive performance of the Gaussian mixture model (GMM) shown in lossy compression, we generate a lossless network architecture with GMM. Besides noticing the successful achievements of attention modules and autoregressive models, we propose to utilize attention modules and add an extra autoregressive model for raw images in our network architecture to boost the performance. Experimental results show that our approach outperforms most classical lossless compression methods and existing learning-based methods.

Shimon Murai, Heming Sun, Jiro Katto

Supported by powerful generative models, low-bitrate learned image compression (LIC) models utilizing perceptual metrics have become feasible. Some of the most advanced models achieve high compression rates and superior perceptual quality by using image captions as sub-information. This paper demonstrates that using a large multi-modal model (LMM), it is possible to generate captions and compress them within a single model. We also propose a novel semantic-perceptual-oriented fine-tuning method applicable to any LIC network, resulting in a 41.58\% improvement in LPIPS BD-rate compared to existing methods. Our implementation and pre-trained weights are available at https://github.com/tokkiwa/ImageTextCoding.

Shota Hirose, Kazuki Kotoyori, Kasidis Arunruangsirilert et al.

Transmission latency significantly affects users' quality of experience in real-time interaction and actuation. As latency is principally inevitable, video prediction can be utilized to mitigate the latency and ultimately enable zero-latency transmission. However, most of the existing video prediction methods are computationally expensive and impractical for real-time applications. In this work, we therefore propose real-time video prediction towards the zero-latency interaction over networks, called IFRVP (Intermediate Feature Refinement Video Prediction). Firstly, we propose three training methods for video prediction that extend frame interpolation models, where we utilize a simple convolution-only frame interpolation network based on IFRNet. Secondly, we introduce ELAN-based residual blocks into the prediction models to improve both inference speed and accuracy. Our evaluations show that our proposed models perform efficiently and achieve the best trade-off between prediction accuracy and computational speed among the existing video prediction methods. A demonstration movie is also provided at http://bit.ly/IFRVPDemo. The code will be released at https://github.com/FykAikawa/IFRVP.

Chao Liu, Heming Sun, Jiro Katto et al.

In this paper, we propose a learned video codec with a residual prediction network (RP-Net) and a feature-aided loop filter (LF-Net). For the RP-Net, we exploit the residual of previous multiple frames to further eliminate the redundancy of the current frame residual. For the LF-Net, the features from residual decoding network and the motion compensation network are used to aid the reconstruction quality. To reduce the complexity, a light ResNet structure is used as the backbone for both RP-Net and LF-Net. Experimental results illustrate that we can save about 10% BD-rate compared with previous learned video compression frameworks. Moreover, we can achieve faster coding speed due to the ResNet backbone. This project is available at https://github.com/chaoliu18/RPLVC.

Xinliang Frederick Zhang, Heming Sun, Xiang Yue et al.

We present a large, challenging dataset, COUGH, for COVID-19 FAQ retrieval. Similar to a standard FAQ dataset, COUGH consists of three parts: FAQ Bank, Query Bank and Relevance Set. The FAQ Bank contains ~16K FAQ items scraped from 55 credible websites (e.g., CDC and WHO). For evaluation, we introduce Query Bank and Relevance Set, where the former contains 1,236 human-paraphrased queries while the latter contains ~32 human-annotated FAQ items for each query. We analyze COUGH by testing different FAQ retrieval models built on top of BM25 and BERT, among which the best model achieves 48.8 under P@5, indicating a great challenge presented by COUGH and encouraging future research for further improvement. Our COUGH dataset is available at https://github.com/sunlab-osu/covid-faq.

Zhibo Chen, Heming Sun, Li Zhang et al.

This paper provides a survey of the latest developments in visual signal coding and processing with generative models. Specifically, our focus is on presenting the advancement of generative models and their influence on research in the domain of visual signal coding and processing. This survey study begins with a brief introduction of well-established generative models, including the Variational Autoencoder (VAE) models, Generative Adversarial Network (GAN) models, Autoregressive (AR) models, Normalizing Flows and Diffusion models. The subsequent section of the paper explores the advancements in visual signal coding based on generative models, as well as the ongoing international standardization activities. In the realm of visual signal processing, our focus lies on the application and development of various generative models in the research of visual signal restoration. We also present the latest developments in generative visual signal synthesis and editing, along with visual signal quality assessment using generative models and quality assessment for generative models. The practical implementation of these studies is closely linked to the investigation of fast optimization. This paper additionally presents the latest advancements in fast optimization on visual signal coding and processing with generative models. We hope to advance this field by providing researchers and practitioners a comprehensive literature review on the topic of visual signal coding and processing with generative models.

Tianhao Peng, Ge Gao, Heming Sun et al.

In recent years, end-to-end learnt video codecs have demonstrated their potential to compete with conventional coding algorithms in term of compression efficiency. However, most learning-based video compression models are associated with high computational complexity and latency, in particular at the decoder side, which limits their deployment in practical applications. In this paper, we present a novel model-agnostic pruning scheme based on gradient decay and adaptive layer-wise distillation. Gradient decay enhances parameter exploration during sparsification whilst preventing runaway sparsity and is superior to the standard Straight-Through Estimation. The adaptive layer-wise distillation regulates the sparse training in various stages based on the distortion of intermediate features. This stage-wise design efficiently updates parameters with minimal computational overhead. The proposed approach has been applied to three popular end-to-end learnt video codecs, FVC, DCVC, and DCVC-HEM. Results confirm that our method yields up to 65% reduction in MACs and 2x speed-up with less than 0.3dB drop in BD-PSNR. Supporting code and supplementary material can be downloaded from: https://jasminepp.github.io/lightweightdvc/

Jinming Liu, Yuntao Wei, Junyan Lin et al.

While learned image compression methods have achieved impressive results in either human visual perception or machine vision tasks, they are often specialized only for one domain. This drawback limits their versatility and generalizability across scenarios and also requires retraining to adapt to new applications-a process that adds significant complexity and cost in real-world scenarios. In this study, we introduce an innovative semantics DISentanglement and COmposition VERsatile codec (DISCOVER) to simultaneously enhance human-eye perception and machine vision tasks. The approach derives a set of labels per task through multimodal large models, which grounding models are then applied for precise localization, enabling a comprehensive understanding and disentanglement of image components at the encoder side. At the decoding stage, a comprehensive reconstruction of the image is achieved by leveraging these encoded components alongside priors from generative models, thereby optimizing performance for both human visual perception and machine-based analytical tasks. Extensive experimental evaluations substantiate the robustness and effectiveness of DISCOVER, demonstrating superior performance in fulfilling the dual objectives of human and machine vision requirements.

Kazuki Kotoyori, Shota Hirose, Heming Sun et al.

Accurate video prediction by deep neural networks, especially for dynamic regions, is a challenging task in computer vision for critical applications such as autonomous driving, remote working, and telemedicine. Due to inherent uncertainties, existing prediction models often struggle with the complexity of motion dynamics and occlusions. In this paper, we propose a novel stochastic long-term video prediction model that focuses on dynamic regions by employing a hybrid warping strategy. By integrating frames generated through forward and backward warpings, our approach effectively compensates for the weaknesses of each technique, improving the prediction accuracy and realism of moving regions in videos while also addressing uncertainty by making stochastic predictions that account for various motions. Furthermore, considering real-time predictions, we introduce a MobileNet-based lightweight architecture into our model. Our model, called SVPHW, achieves state-of-the-art performance on two benchmark datasets.

Ruoyu Feng, Jinming Liu, Xin Jin et al.

Image coding for machines (ICM) aims to compress images to support downstream AI analysis instead of human perception. For ICM, developing a unified codec to reduce information redundancy while empowering the compressed features to support various vision tasks is very important, which inevitably faces two core challenges: 1) How should the compression strategy be adjusted based on the downstream tasks? 2) How to well adapt the compressed features to different downstream tasks? Inspired by recent advances in transferring large-scale pre-trained models to downstream tasks via prompting, in this work, we explore a new ICM framework, termed Prompt-ICM. To address both challenges by carefully learning task-driven prompts to coordinate well the compression process and downstream analysis. Specifically, our method is composed of two core designs: a) compression prompts, which are implemented as importance maps predicted by an information selector, and used to achieve different content-weighted bit allocations during compression according to different downstream tasks; b) task-adaptive prompts, which are instantiated as a few learnable parameters specifically for tuning compressed features for the specific intelligent task. Extensive experiments demonstrate that with a single feature codec and a few extra parameters, our proposed framework could efficiently support different kinds of intelligent tasks with much higher coding efficiency.

Zhao Zan, Chao Liu, Heming Sun et al.

Learned image compression techniques have achieved considerable development in recent years. In this paper, we find that the performance bottleneck lies in the use of a single hyperprior decoder, in which case the ternary Gaussian model collapses to a binary one. To solve this, we propose to use three hyperprior decoders to separate the decoding process of the mixed parameters in discrete Gaussian mixture likelihoods, achieving more accurate parameters estimation. Experimental results demonstrate the proposed method optimized by MS-SSIM achieves on average 3.36% BD-rate reduction compared with state-of-the-art approach. The contribution of the proposed method to the coding time and FLOPs is negligible.

Chao Liu, Heming Sun, Jiro Katto et al.

Convolutional neural network (CNN)-based filters have achieved great success in video coding. However, in most previous works, individual models are needed for each quantization parameter (QP) band. This paper presents a generic method to help an arbitrary CNN-filter handle different quantization noise. We model the quantization noise problem and implement a feasible solution on CNN, which introduces the quantization step (Qstep) into the convolution. When the quantization noise increases, the ability of the CNN-filter to suppress noise improves accordingly. This method can be used directly to replace the (vanilla) convolution layer in any existing CNN-filters. By using only 25% of the parameters, the proposed method achieves better performance than using multiple models with VTM-6.3 anchor. Besides, an additional BD-rate reduction of 0.2% is achieved by our proposed method for chroma components.

Chao Liu, Heming Sun, Jiro Katto et al.

Convolutional Neural Network (CNN)-based filters have achieved significant performance in video artifacts reduction. However, the high complexity of existing methods makes it difficult to be applied in real usage. In this paper, a CNN-based low complexity filter is proposed. We utilize depth separable convolution (DSC) merged with the batch normalization (BN) as the backbone of our proposed CNN-based network. Besides, a weight initialization method is proposed to enhance the training performance. To solve the well known over smoothing problem for the inter frames, a frame-level residual mapping (RM) is presented. We analyze some of the mainstream methods like frame-level and block-level based filters quantitatively and build our CNN-based filter with frame-level control to avoid the extra complexity and artificial boundaries caused by block-level control. In addition, a novel module called RM is designed to restore the distortion from the learned residuals. As a result, we can effectively improve the generalization ability of the learning-based filter and reach an adaptive filtering effect. Moreover, this module is flexible and can be combined with other learning-based filters. The experimental results show that our proposed method achieves significant BD-rate reduction than H.265/HEVC. It achieves about 1.2% BD-rate reduction and 79.1% decrease in FLOPs than VR-CNN. Finally, the measurement on H.266/VVC and ablation studies are also conducted to ensure the effectiveness of the proposed method.

Chao Liu, Heming Sun, Junan Chen et al.

In this paper, a dual learning-based method in intra coding is introduced for PCS Grand Challenge. This method is mainly composed of two parts: intra prediction and reconstruction filtering. They use different network structures, the neural network-based intra prediction uses the full-connected network to predict the block while the neural network-based reconstruction filtering utilizes the convolutional networks. Different with the previous filtering works, we use a network with more powerful feature extraction capabilities in our reconstruction filtering network. And the filtering unit is the block-level so as to achieve a more accurate filtering compensation. To our best knowledge, among all the learning-based methods, this is the first attempt to combine two different networks in one application, and we achieve the state-of-the-art performance for AI configuration on the HEVC Test sequences. The experimental result shows that our method leads to significant BD-rate saving for provided 8 sequences compared to HM-16.20 baseline (average 10.24% and 3.57% bitrate reductions for all-intra and random-access coding, respectively). For HEVC test sequences, our model also achieved a 9.70% BD-rate saving compared to HM-16.20 baseline for all-intra configuration.

Zhengxue Cheng, Heming Sun, Masaru Takeuchi et al.

Image compression has been investigated as a fundamental research topic for many decades. Recently, deep learning has achieved great success in many computer vision tasks, and is gradually being used in image compression. In this paper, we present a lossy image compression architecture, which utilizes the advantages of convolutional autoencoder (CAE) to achieve a high coding efficiency. First, we design a novel CAE architecture to replace the conventional transforms and train this CAE using a rate-distortion loss function. Second, to generate a more energy-compact representation, we utilize the principal components analysis (PCA) to rotate the feature maps produced by the CAE, and then apply the quantization and entropy coder to generate the codes. Experimental results demonstrate that our method outperforms traditional image coding algorithms, by achieving a 13.7% BD-rate decrement on the Kodak database images compared to JPEG2000. Besides, our method maintains a moderate complexity similar to JPEG2000.