Parameter-Efficient Tuning (PET) has been widely studied to adapt pretrained vision-language models (VLMs) to various downstream tasks. Most existing PET methods are limited to uni-modal tuning of textual modality and cannot fully exploit the generalization ability of VLMs. Recent multi-modal tuning methods cannot address the catastrophic performance loss in generalizing to unseen classes without incurring significant memory and computation overheads. In this paper, we reveal the multi-modal forgetting issue that violates model coupling for constraining visual features and textual embeddings in simultaneously tuning both modalities, and propose a novel Vision Language Efficient Tuning (VioLET) framework to address the issue. We first propose VioLET-CMG that leverages Collaborative Multi-modal Gradients (CMG) by incorporating an additional visual encoder to provide extra textual gradients to regulate the multi-modal tuning. VioLET-CMG avoids conflicts between visual and textual gradients and guarantee modal coupling with textual gradient orthogonalization. Furthermore, we present VioLET-CarE that achieves quasi-orthogonal textual tuning by Controlling language hyperspherical Energy (CarE) to remove the additional visual encoder for enabling collaborative multi-modal gradients in VioLET-CMG. VioLET-CarE enhances computational efficiency via additive low-rank feature transformation for textual tuning. Experimental results demonstrate that the proposed VioLET framework consistently achieves state-of-the-art performance in new class generalization and few-shot recognition using both ResNet-50 and ViT/B-16 backbones. Remarkably, compared with VioLET-CMG, VioLET-CarE further reduces GPU memory cost by 27

Existing video frame interpolation (VFI) methods often adopt a frame-centric approach, processing videos as independent short segments (e.g., triplets), which leads to temporal inconsistencies and motion artifacts. To overcome this, we propose a holistic, video-centric paradigm named \textbf{L}ocal \textbf{D}iffusion \textbf{F}orcing for \textbf{V}ideo \textbf{F}rame \textbf{I}nterpolation (LDF-VFI). Our framework is built upon an auto-regressive diffusion transformer that models the entire video sequence to ensure long-range temporal coherence. To mitigate error accumulation inherent in auto-regressive generation, we introduce a novel skip-concatenate sampling strategy that effectively maintains temporal stability. Furthermore, LDF-VFI incorporates sparse, local attention and tiled VAE encoding, a combination that not only enables efficient processing of long sequences but also allows generalization to arbitrary spatial resolutions (e.g., 4K) at inference without retraining. An enhanced conditional VAE decoder, which leverages multi-scale features from the input video, further improves reconstruction fidelity. Empirically, LDF-VFI achieves state-of-the-art performance on challenging long-sequence benchmarks, demonstrating superior per-frame quality and temporal consistency, especially in scenes with large motion.

Point cloud attribute compression is challenged by fitting the attribute signals living on irregular geometric structures. Existing methods cannot achieve compact multiscale representation for high-fidelity reconstruction using the handcrafted transforms or deep learning-based techniques. In this paper, we propose a novel geometry-aware lifting-based multiscale network via spatial-channel lifting scheme for point cloud attribute compression. The proposed network cascades geometry-aware spatial lifting to reduce spatial redundancy by adaptively capturing irregular geometric structures and progressive channel lifting to progressively reduce channel-wise redundancy in multiscale representation. Furthermore, we design the split, predict, and update operations for geometry-aware spatial lifting to fully exploit the geometry information representing irregular structures. We develop geometry-aware adaptive split to equally split input points with significance scores indicating their dependencies, and propose geometry-aware cross-attention filtering for the predict and update operations for decorrelation based on geometry information. To our best knowledge, this paper achieves the first lifting-based learned transform for point cloud compression that enjoys reversibility guarantees of multiscale representation to enhance rate-distortion performance. Experimental results show that the proposed framework achieves state-of-the-art performance on extensive point cloud datasets, and outperforms latest MPEG G-PCC standard and most recent deep learning based methods.

We propose Noise Conditional Variational Score Distillation (NCVSD), a novel method for distilling pretrained diffusion models into generative denoisers. We achieve this by revealing that the unconditional score function implicitly characterizes the score function of denoising posterior distributions. By integrating this insight into the Variational Score Distillation (VSD) framework, we enable scalable learning of generative denoisers capable of approximating samples from the denoising posterior distribution across a wide range of noise levels. The proposed generative denoisers exhibit desirable properties that allow fast generation while preserve the benefit of iterative refinement: (1) fast one-step generation through sampling from pure Gaussian noise at high noise levels; (2) improved sample quality by scaling the test-time compute with multi-step sampling; and (3) zero-shot probabilistic inference for flexible and controllable sampling. We evaluate NCVSD through extensive experiments, including class-conditional image generation and inverse problem solving. By scaling the test-time compute, our method outperforms teacher diffusion models and is on par with consistency models of larger sizes. Additionally, with significantly fewer NFEs than diffusion-based methods, we achieve record-breaking LPIPS on inverse problems.

Existing frameworks for learned video compression suffer from a dilemma between inaccurate temporal alignment and error propagation for motion estimation and compensation (ME/MC). The separate-transform framework employs distinct transforms for intra-frame and inter-frame compression to yield impressive rate-distortion (R-D) performance but causes evident error propagation, while the unified-transform framework eliminates error propagation via shared transforms but is inferior in ME/MC in shared latent domains. To address this limitation, in this paper, we propose a novel unifiedtransform framework with dual-domain progressive temporal alignment and quality-conditioned mixture-of-expert (QCMoE) to enable quality-consistent and error-propagation-free streaming for learned video compression. Specifically, we propose dualdomain progressive temporal alignment for ME/MC that leverages coarse pixel-domain alignment and refined latent-domain alignment to significantly enhance temporal context modeling in a coarse-to-fine fashion. The coarse pixel-domain alignment efficiently handles simple motion patterns with optical flow estimated from a single reference frame, while the refined latent-domain alignment develops a Flow-Guided Deformable Transformer (FGDT) over latents from multiple reference frames to achieve long-term motion refinement (LTMR) for complex motion patterns. Furthermore, we design a QCMoE module for continuous bit-rate adaptation that dynamically assigns different experts to adjust quantization steps per pixel based on target quality and content rather than relies on a single quantization step. QCMoE allows continuous and consistent rate control with appealing R-D performance. Experimental results show that the proposed method achieves competitive R-D performance compared with the state-of-the-arts, while successfully eliminating error propagation.

Monocular 3D human pose estimation (HPE) methods estimate the 3D positions of joints from individual images. Existing 3D HPE approaches often use the cropped image alone as input for their models. However, the relative depths of joints cannot be accurately estimated from cropped images without the corresponding camera intrinsics, which determine the perspective relationship between 3D objects and the cropped images. In this work, we introduce Perspective Encoding (PE) to encode the camera intrinsics of the cropped images. Moreover, since the human subject can appear anywhere within the original image, the perspective relationship between the 3D scene and the cropped image differs significantly, which complicates model fitting. Additionally, the further the human subject deviates from the image center, the greater the perspective distortions in the cropped image. To address these issues, we propose Perspective Rotation (PR), a transformation applied to the original image that centers the human subject, thereby reducing perspective distortions and alleviating the difficulty of model fitting. By incorporating PE and PR, we propose a novel 3D HPE framework, PersPose. Experimental results demonstrate that PersPose achieves state-of-the-art (SOTA) performance on the 3DPW, MPIINF-3DHP, and Human3.6M datasets. For example, on the in-the-wild dataset 3DPW, PersPose achieves an MPJPE of 60.1 mm, 7.54

Existing 2D-to-3D human pose estimation (HPE) methods struggle with the occlusion issue by enriching information like temporal and visual cues in the lifting stage. In this paper, we argue that these methods ignore the limitation of the sparse skeleton 2D input representation, which fundamentally restricts the 2D-to-3D lifting and worsens the occlusion issue. To address these, we propose a novel two-stage generative densification method, named Hierarchical Pose AutoRegressive Transformer (HiPART), to generate hierarchical 2D dense poses from the original sparse 2D pose. Specifically, we first develop a multi-scale skeleton tokenization module to quantize the highly dense 2D pose into hierarchical tokens and propose a Skeleton-aware Alignment to strengthen token connections. We then develop a Hierarchical AutoRegressive Modeling scheme for hierarchical 2D pose generation. With generated hierarchical poses as inputs for 2D-to-3D lifting, the proposed method shows strong robustness in occluded scenarios and achieves state-of-the-art performance on the single-frame-based 3D HPE. Moreover, it outperforms numerous multi-frame methods while reducing parameter and computational complexity and can also complement them to further enhance performance and robustness.

The aging of operational reactors leads to increased mechanical vibrations in the reactor interior.The vibration of the in-core sensors near their nominal locations is a new problem for neutronic field reconstruction.Current field-reconstruction methods fail to handle spatially moving sensors.In this study,we propose a Voronoi tessellation technique in combination with convolutional neural networks to handle this challenge.Observations from movable in-core sensors were projected onto the same global field structure using Voronoi tessellation,holding the magnitude and location information of the sensors.General convolutional neural networks were used to learn maps from observations to the global field.The proposed method reconstructed multi-physics fields(including fast flux,thermal flux,and power rate)using observations from a single field(such as thermal flux).Numerical tests based on the IAEA benchmark demonstrated the potential of the proposed method in practical engineering applications,particularly within an amplitude of 5 cm around the nominal locations,which led to average relative errors below 5％and 10％in the L2 and L∞norms,respectively.

Learned image compression (LIC) has gained traction as an effective solution for image storage and transmission in recent years. However, existing LIC methods are redundant in latent representation due to limitations in capturing anisotropic frequency components and preserving directional details. To overcome these challenges, we propose a novel frequency-aware transformer (FAT) block that for the first time achieves multiscale directional ananlysis for LIC. The FAT block comprises frequency-decomposition window attention (FDWA) modules to capture multiscale and directional frequency components of natural images. Additionally, we introduce frequency-modulation feed-forward network (FMFFN) to adaptively modulate different frequency components, improving rate-distortion performance. Furthermore, we present a transformer-based channel-wise autoregressive (T-CA) model that effectively exploits channel dependencies. Experiments show that our method achieves state-of-the-art rate-distortion performance compared to existing LIC methods, and evidently outperforms latest standardized codec VTM-12.1 by 14.5%, 15.1%, 13.0% in BD-rate on the Kodak, Tecnick, and CLIC datasets.

Image compression for machine and human vision (ICMH) has gained increasing attention in recent years. Existing ICMH methods are limited by high training and storage overheads due to heavy design of task-specific networks. To address this issue, in this paper, we develop a novel lightweight adapter-based tuning framework for ICMH, named Adapt-ICMH, that better balances task performance and bitrates with reduced overheads. We propose a spatial-frequency modulation adapter (SFMA) that simultaneously eliminates non-semantic redundancy with a spatial modulation adapter, and enhances task-relevant frequency components and suppresses task-irrelevant frequency components with a frequency modulation adapter. The proposed adapter is plug-and-play and compatible with almost all existing learned image compression models without compromising the performance of pre-trained models. Experiments demonstrate that Adapt-ICMH consistently outperforms existing ICMH frameworks on various machine vision tasks with fewer fine-tuned parameters and reduced computational complexity. Code will be released at https://github.com/qingshi9974/ECCV2024-AdpatICMH.

Data mixing methods play a crucial role in semi-supervised learning (SSL), but their application is unexplored in long-tailed semi-supervised learning (LTSSL). The primary reason is that the in-batch mixing manner fails to address class imbalance. Furthermore, existing LTSSL methods mainly focus on re-balancing data quantity but ignore class-wise uncertainty, which is also vital for class balance. For instance, some classes with sufficient samples might still exhibit high uncertainty due to indistinguishable features. To this end, this paper introduces the Balanced and Entropy-based Mix (BEM), a pioneering mixing approach to re-balance the class distribution of both data quantity and uncertainty. Specifically, we first propose a class balanced mix bank to store data of each class for mixing. This bank samples data based on the estimated quantity distribution, thus re-balancing data quantity. Then, we present an entropy-based learning approach to re-balance class-wise uncertainty, including entropy-based sampling strategy, entropy-based selection module, and entropy-based class balanced loss. Our BEM first leverages data mixing for improving LTSSL, and it can also serve as a complement to the existing re-balancing methods. Experimental results show that BEM significantly enhances various LTSSL frameworks and achieves state-of-the-art performances across multiple benchmarks.

As an interesting technique that allows data extraction and image recovery without any loss, reversible data hiding in encrypted images is an area of great concern in the field of information security. In this paper, a new reversible data hiding method with high embedding capacity is proposed based on hybrid prediction and Huffman coding. The combination of two embedding mechanisms is innovatively designed to improve the embedding capacity according to different parts of the original image, i.e., the most significant bit-plane and the remaining seven bit-planes. In the first part of this method, the prediction value of each pixel is obtained by calculating the average value of its two neighboring pixels, and all of the most significant bits can be vacated to accommodate additional data. In the second part, the prediction value of each pixel is calculated using the median edge detector predictor, on which the tag map is built. Then, Huffman coding is used to compress the tag map so that a large amount of vacant space is obtained. Finally, the secret data can be embedded into the vacated space by directly using bit substitution. Compared with some recently reported methods, experimental results and analysis have demonstrated that an original image with high visual symmetry/quality can be recovered. Also, larger embedding capacity can be achieved, such as 3.3894 bpp and 3.2824 bpp, for BOSSBase and BOWS2 databases, respectively.

There has been a recent surge of interest in introducing transformers to 3D human pose estimation (HPE) due to their powerful capabilities in modeling long-term dependencies. However, existing transformer-based methods treat body joints as equally important inputs and ignore the prior knowledge of human skeleton topology in the self-attention mechanism. To tackle this issue, in this paper, we propose a Pose-Oriented Transformer (POT) with uncertainty guided refinement for 3D HPE. Specifically, we first develop novel pose-oriented self-attention mechanism and distance-related position embedding for POT to explicitly exploit the human skeleton topology. The pose-oriented self-attention mechanism explicitly models the topological interactions between body joints, whereas the distance-related position embedding encodes the distance of joints to the root joint to distinguish groups of joints with different difficulties in regression. Furthermore, we present an Uncertainty-Guided Refinement Network (UGRN) to refine pose predictions from POT, especially for the difficult joints, by considering the estimated uncertainty of each joint with uncertainty-guided sampling strategy and self-attention mechanism. Extensive experiments demonstrate that our method significantly outperforms the state-of-the-art methods with reduced model parameters on 3D HPE benchmarks such as Human3.6M and MPI-INF-3DHP.

Recent 2D-to-3D human pose estimation (HPE) utilizes temporal consistency across sequences to alleviate the depth ambiguity problem but ignore the action related prior knowledge hidden in the pose sequence. In this paper, we propose a plug-and-play module named Action Prompt Module (APM) that effectively mines different kinds of action clues for 3D HPE. The highlight is that, the mining scheme of APM can be widely adapted to different frameworks and bring consistent benefits. Specifically, we first present a novel Action-related Text Prompt module (ATP) that directly embeds action labels and transfers the rich language information in the label to the pose sequence. Besides, we further introduce Action-specific Pose Prompt module (APP) to mine the position-aware pose pattern of each action, and exploit the correlation between the mined patterns and input pose sequence for further pose refinement. Experiments show that APM can improve the performance of most video-based 2D-to-3D HPE frameworks by a large margin.

As the rapid development of third-party storage and homomorphic encryption have profoundly stimulated the desire for secure communication, reversible data hiding in encrypted images has received widespread attention, since it allows lossless data conveying and perfect image recovery. In order to obtain secure reversible data hiding with high embedding capacity, a novel block embedding method is proposed, based on an adaptive recognition strategy for combined blocks in the binary image, with which the adjacent identical blocks can be integrated into a combination to reserve more spare bits for data accommodation. Furthermore, a fully reversible data hiding method for grayscale images in the encryption domain is designed. The secret data is hidden into lower bit-planes of the image while the original bits of those embedded lower pixels are recorded into the vacated space of higher bit-planes. The original image can be reconstructed flawlessly as well as the secret data being extracted without errors. To reinforce security, the original image and the secret data are encrypted and scrambled based on sequences generated with the high-dimension chaotic system. Due to its high sensitivity of initial values, the performance such as security and robustness is guaranteed. By comparing the PSNR value of the marked decrypted image and evaluating the quality of the extracted secret image, experimental results demonstrate that the proposed method can obtain higher embedding capacity, achieving 0.2700–0.3924 bpp increment over the state-of-the-art methods, and recover the marked decrypted image with high visual symmetry/quality, and efficiently resist against potential attacks, such as the histogram analysis, differential, brute-force, JPEG attacks, and so on.

Transformers have recently shown superior performance than CNN on semantic segmentation. However, previous works mostly focus on the deliberate design of the encoder, while seldom considering the decoder part. In this paper, we find that a light weighted decoder counts for segmentation, and propose a pure transformer-based segmentation decoder, named SegDeformer, to seamlessly incorporate into current varied transformer-based encoders. The highlight is that SegDeformer is able to conveniently utilize the tokenized input and the attention mechanism of the transformer for effective context mining. This is achieved by two key component designs, i.e., the internal and external context mining modules. The former is equipped with internal attention within an image to better capture global-local context, while the latter introduces external tokens from other images to enhance current representation. To enable SegDeformer in a scalable way, we further provide performance/efficiency optimization modules for flexible deployment. Experiments on widely used benchmarks ADE20K, COCO-Stuff and Cityscapes and different transformer encoders (e.g., ViT, MiT and Swin) demonstrate that SegDeformer can bring consistent performance gains. Code is available at https://github.com/lygsbw/segdeformer .

Recent 2D-to-3D human pose estimation works tend to utilize the graph structure formed by the topology of the human skeleton. However, we argue that this skeletal topology is too sparse to reflect the body structure and suffer from serious 2D-to-3D ambiguity problem. To overcome these weaknesses, we propose a novel graph convolution network architecture, Hierarchical Graph Networks (HGN). It is based on denser graph topology generated by our multi-scale graph structure building strategy, thus providing more delicate geometric information. The proposed architecture contains three sparse-to-fine representation subnetworks organized in parallel, in which multi-scale graph-structured features are processed and exchange information through a novel feature fusion strategy, leading to rich hierarchical representations. We also introduce a 3D coarse mesh constraint to further boost detail-related feature learning. Extensive experiments demonstrate that our HGN achieves the state-of-the art performance with reduced network parameters. Code is released at https://github.com/qingshi9974/BMVC2021-Hierarchical-Graph-Networks-for-3D-Human-Pose-Estimation.

针对现有三维形状配准方法中存在左右翻转的错误匹配问题,提出了基于内蕴对称特征检测的高效形状配准算法.首先,通过热核与几何约束构建模型的内蕴自对称点对;其次,基于谱嵌入特征空间分析提取模型的内蕴对称平面,并依据模型表面法向量有效识别模型的左右结构属性;然后,根据内蕴对称点对获得模型的一致性谱对称结构描述;最后,引入一致性点漂移算法(CPD),实现基于谱对称的非刚性模型的形状配准,有效避免了模型配准中的左右结构翻转问题.实验进一步论证了这种方法不仅有效提高了模型匹配的效率,而且能有效识别同类模型的结构特征,对于非刚性模型的配准具有较强的鲁棒性.

针对传统的三维模型单一特征识别精度低问题, 提出一种自适应稀疏编码融合的非刚性三维模型分类算法.首先提取模型的平均测地线特征、热核特征、形状直径函数特征, 以构造互补的多特征形状描述; 其次利用特征袋模型构造AGD-BoF, HKS-BoF, SDF-BoF特征向量, 通过随机样本建立特征权值矩阵; 最后利用特征权值矩阵与稀疏优化编码自适应融合, 并采用 Softmax 分类算法实现非刚性模型的有效分类. 通过在非刚性数据集 SHREC10 与SHREC11上的综合实验表明, 自适应稀疏编码融合的分类算法具有更高的识别准确率及较强的鲁棒性.