Large Language Models (LLMs) have revolutionized natural language processing tasks. However, their practical application is constrained by substantial memory and computational demands. Post-training quantization (PTQ) is considered an effective method to accelerate LLM inference. Despite its growing popularity in LLM model compression, PTQ deployment faces two major challenges. First, low-bit quantization leads to performance degradation. Second, restricted by the limited integer computing unit type on GPUs, quantized matrix operations with different precisions cannot be effectively accelerated. To address these issues, we introduce a novel arbitrary-bit quantization algorithm and inference framework, ABQ-LLM. It achieves superior performance across various quantization settings and enables efficient arbitrary-precision quantized inference on the GPU. ABQ-LLM introduces several key innovations: (1) a distribution correction method for transformer blocks to mitigate distribution differences caused by full quantization of weights and activations, improving performance at low bit-widths. (2) the bit balance strategy to counteract performance degradation from asymmetric distribution issues at very low bit-widths (e.g., 2-bit). (3) an innovative quantization acceleration framework that reconstructs the quantization matrix multiplication of arbitrary precision combinations based on BTC (Binary TensorCore) equivalents, gets rid of the limitations of INT4/INT8 computing units. ABQ-LLM can convert each component bit width gain into actual acceleration gain, maximizing performance under mixed precision(e.g., W6A6, W2A8). Based on W2*A8 quantization configuration on LLaMA-7B model, it achieved a WikiText2 perplexity of 7.59 (2.17⬇ vs 9.76 in AffineQuant). Compared to SmoothQuant, we realized 1.6x acceleration improvement and 2.7x memory compression gain.

In this paper, we present a new representation for neural radiance fields that accelerates both the training and the inference processes with VDB, a hierarchical data structure for sparse volumes. VDB takes both the advantages of sparse and dense volumes for compact data representation and efficient data access, being a promising data structure for NeRF data interpolation and ray marching. Our method, Plenoptic VDB (PlenVDB), directly learns the VDB data structure from a set of posed images by means of a novel training strategy and then uses it for real-time rendering. Experimental results demonstrate the effectiveness and the efficiency of our method over previous arts: First, it converges faster in the training process. Second, it delivers a more compact data format for NeRF data presentation. Finally, it renders more efficiently on commodity graphics hardware. Our mobile PlenVDB demo achieves 30+ FPS, 1280×720 resolution on an iPhone12 mobile phone. Check plenvdb.github.io for details.

In this work, we propose a novel meta-learning approach for few-shot classification, which learns transferable prior knowledge across tasks and directly produces network parameters for similar unseen tasks with training samples. Our approach, called LGM-Net, includes two key modules, namely, TargetNet and MetaNet. The TargetNet module is a neural network for solving a specific task and the MetaNet module aims at learning to generate functional weights for TargetNet by observing training samples. We also present an intertask normalization strategy for the training process to leverage common information shared across different tasks. The experimental results on Omniglot and miniImageNet datasets demonstrate that LGM-Net can effectively adapt to similar unseen tasks and achieve competitive performance, and the results on synthetic datasets show that transferable prior knowledge is learned by the MetaNet module via mapping training data to functional weights. LGM-Net enables fast learning and adaptation since no further tuning steps are required compared to other meta-learning approaches

Aggregation structures with explicit information, such as image attributes and scene semantics, are effective and popular for intelligent systems for assessing aesthetics of visual data. However, useful information may not be available due to the high cost of manual annotation and expert design. In this paper, we present a novel multi-patch (MP) aggregation method for image aesthetic assessment. Different from state-of-the-art methods, which augment an MP aggregation network with various visual attributes, we train the model in an end-to-end manner with aesthetic labels only (i.e., aesthetically positive or negative). We achieve the goal by resorting to an attention-based mechanism that adaptively adjusts the weight of each patch during the training process to improve learning efficiency. In addition, we propose a set of objectives with three typical attention mechanisms (i.e., average, minimum, and adaptive) and evaluate their effectiveness on the Aesthetic Visual Analysis (AVA) benchmark. Numerical results show that our approach outperforms existing methods by a large margin. We further verify the effectiveness of the proposed attention-based objectives via ablation studies and shed light on the design of aesthetic assessment systems.

In this paper, we present a method for reconstructing the drawing process of Chinese brush paintings. We demonstrate the possibility of computing an artistically reasonable drawing order from a static brush painting that is consistent with the rules of art. We map the key principles of drawing composition to our computational framework, which first organizes the strokes in three stages and then optimizes stroke ordering with natural evolution strategies. Our system produces reasonable animated constructions of Chinese brush paintings with minimal or no user intervention. We test our algorithm on a range of input paintings with varying degrees of complexity and structure and then evaluate the results via a user study. We discuss the applications of the proposed system to painting instruction, painting animation, and image stylization, especially in the context of art teaching.

This paper presents a data-driven approach for automatically generating cartoon faces in different styles from a given portrait image. Our stylization pipeline consists of two steps: an offline analysis step to learn about how to select and compose facial components from the databases; a runtime synthesis step to generate the cartoon face by assembling parts from a database of stylized facial components. We propose an optimization framework that, for a given artistic style, simultaneously considers the desired image-cartoon relationships of the facial components and a proper adjustment of the image composition. We measure the similarity between facial components of the input image and our cartoon database via image feature matching, and introduce a probabilistic framework for modeling the relationships between cartoon facial components. We incorporate prior knowledge about image-cartoon relationships and the optimal composition of facial components extracted from a set of cartoon faces to maintain a natural, consistent, and attractive look of the results. We demonstrate generality and robustness of our approach by applying it to a variety of portrait images and compare our output with stylized results created by artists via a comprehensive user study.

Unsupervised ranking faces one critical challenge in evaluation applications, that is, no ground truth is available. While PageRank and its variants show a good solution in related objects, they are applicable only for ranking from link-structure data. In this work, we focus on unsupervised ranking from multi-attribute data which is also common in evaluation tasks. To overcome the challenge, we propose five essential meta-rules for the design and assessment of unsupervised ranking approaches: scale and translation invariance, strict monotonicity, compatibility of linearity and nonlinearity, smoothness, and explicitness of parameter size. These meta-rules are regarded as high level knowledge for unsupervised ranking tasks. Inspired by the works in [2] and [6], we propose a ranking principal curve (RPC) model, which learns a one-dimensional manifold function to perform unsupervised ranking tasks on multi-attribute observations. Furthermore, the RPC is modeled to be a cubic Bézier curve with control points restricted in the interior of a hypercube, complying with all the five meta-rules to infer a reasonable ranking list. With control points as model parameters, one is able to understand the learned manifold and to interpret and visualize the ranking results. Numerical experiments of the presented RPC model are conducted on two open datasets of different ranking applications. In comparison with the state-of-the-art approaches, the new model is able to show more reasonable ranking lists.

Real-world images usually contain vivid contents and rich textural details, which will complicate the manipulation on them. In this paper, we design a new framework based on exampled-based texture synthesis to enhance content-aware image retargeting. By detecting the textural regions in an image, the textural image content can be synthesized rather than simply distorted or cropped. This method enables the manipulation of textural & non-textural regions with different strategies since they have different natures. We propose to retarget the textural regions by example-based synthesis and non-textural regions by fast multi-operator. To achieve practical retargeting applications for general images, we develop an automatic and fast texture detection method that can detect multiple disjoint textural regions. We adjust the saliency of the image according to the features of the textural regions. To validate the proposed method, comparisons with state-of-the-art image retargeting techniques and a user study were conducted. Convincing visual results are shown to demonstrate the effectiveness of the proposed method.

Similar objects are ubiquitous and abundant in both natural and artificial scenes. Determining the visual importance of several similar objects in a complex photograph is a challenge for image understanding algorithms. This study aims to define the importance of similar objects in an image and to develop a method that can select the most important instances for an input image from multiple similar objects. This task is challenging because multiple objects must be compared without adequate semantic information. This challenge is addressed by building an image database and designing an interactive system to measure object importance from human observers. This ground truth is used to define a range of features related to the visual importance of similar objects. Then, these features are used in learning-to-rank and random forest to rank similar objects in an image. Importance predictions were validated on 5,922 objects. The most important objects can be identified automatically. The factors related to composition (e.g., size, location, and overlap) are particularly informative, although clarity and color contrast are also important. We demonstrate the usefulness of similar object importance on various applications, including image retargeting, image compression, image re-attentionizing, image admixture, and manipulation of blindness images.

Marginal histograms provide valuable information for various computer vision problems. However, current image restoration methods do not fully exploit the potential of marginal histograms, in particular, their role as ensemble constraints on the marginal statistics of the restored image. In this paper, we introduce a new framework, UniHIST, to incorporate marginal histogram constraints into image restoration. The key idea of UniHIST is to minimize the discrepancy between the marginal histograms of the restored image and the reference histograms in pixel or gradient domains using the quadratic Wasserstein (W2) distance. The W2distance can be computed directly from data without resorting to density estimation. It provides a differentiable metric between marginal histograms and allows easy integration with existing image restoration methods. We demonstrate the effectiveness of UniHIST through denoising of pattern images and non-blind deconvolution of natural images. We show that UniHIST enhances restoration performance and leads to visual and quantitative improvements over existing state-of-the-art methods.

Several important classes of images such as text, barcode and pattern images have the property that pixels can only take a distinct subset of values. This knowledge can benefit the restoration of such images, but it has not been widely considered in current restoration methods. In this work, we describe an effective and efficient approach to incorporate the knowledge of distinct pixel values of the pristine images into the general regularized least squares restoration framework. We introduce a new regularizer that attains zero at the designated pixel values and becomes a quadratic penalty function in the intervals between them. When incorporated into the regularized least squares restoration framework, this regularizer leads to a simple and efficient step that resembles and extends the rounding operation, which we term as soft-rounding. We apply the soft-rounding enhanced solution to the restoration of binary text/barcode images and pattern images with multiple distinct pixel values. Experimental results show that soft-rounding enhanced restoration methods achieve significant improvement in both visual quality and quantitative measures (PSNR and SSIM). Furthermore, we show that this regularizer can also benefit the restoration of general natural images.

The study of face alignment has been an area of intense research in computer vision, with its achievements widely used in computer graphics applications. The performance of various face alignment methods is often image-dependent or somewhat random because of their own strategy. This study aims to develop a method that can select an input image with good face alignment results from many results produced by a single method or multiple ones. The task is challenging because different face alignment results need to be evaluated without any ground truth. This study addresses this problem by designing a feasible feature extraction scheme to measure the quality of face alignment results. The feature is then used in various machine learning algorithms to rank different face alignment results. Our experiments show that our method is promising for ranking face alignment results and is able to pick good face alignment results, which can enhance the overall performance of a face alignment method with a random strategy. We demonstrate the usefulness of our ranking-enhanced face alignment algorithm in two practical applications: face cartoon stylization and digital face makeup.

We propose a fast minimax path-based depth interpolation method. The algorithm computes for each target pixel varying contributions from reliable depth seeds, and weighted averaging is used to interpolate missing depths. Compared with state-of-the-art joint geodesic upsampling method which selects the K nearest seeds to interpolate missing depths with O(Kn) complexity, our method does not need to limit the number of seeds to K and reduces the computational complexity to O(n). In addition, the minimax path chooses a path with the smallest maximum immediate pairwise pixel difference on it, so it tends to preserve sharp depth discontinuities better. In contrast to the results of previous depth upsampling algorithms, our approach can provide accurate depths with fewer artifacts.

Unsupervised ranking faces one critical challenge in evaluation applications, that is, no ground truth is available. When PageRank and its variants show a good solution in related objects, they are applicable only for ranking from link-structure data. In this work, we focus on unsupervised ranking from multi-attribute data which is also common in evaluation tasks. To overcome the challenge, we propose five essential meta-rules for the design and assessment of unsupervised ranking approaches: scale and translation invariance, strict monotonicity, compatibility of linearity and nonlinearity, smoothness, and explicitness of parameter size. These meta-rules are regarded as high level knowledge for unsupervised ranking tasks. Inspired by the works in [12] and [35], we propose a ranking principal curve (RPC) model, which learns a one-dimensional manifold function to perform unsupervised ranking tasks on multi-attribute observations. Furthermore, the RPC is modeled to be a cubic Bezier curve with control points restricted in the interior of a hypercube, complying with all the five meta-rules to infer a reasonable ranking list. With control points as model parameters, one is able to understand the learned manifold and to interpret and visualize the ranking results. Numerical experiments of the presented RPC model are conducted on two open datasets of different ranking applications. In comparison with the state-of-the-art approaches, the new model is able to show more reasonable ranking lists.

This paper presents a new non-blind image restoration method based on the symmetric generalized Pareto (SGP) prior, which models the heavy-tailed distributions of gradients for natural images. Through experiments we show that the SGP model achieves log likelihood scores comparable to the hyper-Laplacian model when fitted to gradients and other band-pass filter responses. More importantly, when incorporated into a Bayesian MAP framework for non-blind image restoration, the SGP model leads to a closed-form solution for a per-pixel subproblem, which affords computational advantages in comparison with the numerical solutions induced from the hyper-Laplacian model. Experimental results show that our method is comparable to existing methods in restoration quality and processing speed.

Integral imaging display (IID) is a promising technology to provide realistic 3D image without glasses. To achieve a large screen IID with a reasonable fabrication cost, a potential solution is a tiled-lens-array IID (TLA-IID). However, TLA-IIDs are subject to 3D image artifacts when there are even slight misalignments between the lens arrays. This work aims at compensating these artifacts by calibrating the lens array poses with a camera and including them in a ray model used for rendering the 3D image. Since the lens arrays are transparent, this task is challenging for traditional calibration methods. In this paper, we propose a novel calibration method based on defining a set of principle observation rays that pass lens centers of the TLA and the camera's optical center. The method is able to determine the lens array poses with only one camera at an arbitrary unknown position without using any additional markers. The principle observation rays are automatically extracted using a structured light based method from a dense correspondence map between the displayed and captured pixels. Experiments show that lens array misalignments can be estimated with a standard deviation smaller than 0.4 pixels. Based on this, 3D image artifacts are shown to be effectively removed in a test TLA-IID with challenging misalignments.

This paper presents a novel tree-based cost aggregation method for dense stereo matching. Instead of employing the minimum spanning tree (MST) and its variants, a new tree structure, "Segment-Tree", is proposed for non-local matching cost aggregation. Conceptually, the segment-tree is constructed in a three-step process: first, the pixels are grouped into a set of segments with the reference color or intensity image, second, a tree graph is created for each segment, and in the final step, these independent segment graphs are linked to form the segment-tree structure. In practice, this tree can be efficiently built in time nearly linear to the number of the image pixels. Compared to MST where the graph connectivity is determined with local edge weights, our method introduces some 'non-local' decision rules: the pixels in one perceptually consistent segment are more likely to share similar disparities, and therefore their connectivity within the segment should be first enforced in the tree construction process. The matching costs are then aggregated over the tree within two passes. Performance evaluation on 19 Middlebury data sets shows that the proposed method is comparable to previous state-of-the-art aggregation methods in disparity accuracy and processing speed. Furthermore, the tree structure can be refined with the estimated disparities, which leads to consistent scene segmentation and significantly better aggregation results.

Abstract This paper presents a novel content‐based method for transferring the colour patterns between images. Unlike previous methods that rely on image colour statistics, our method puts an emphasis on high‐level scene content analysis. We first automatically extract the foreground subject areas and background scene layout from the scene. The semantic correspondences of the regions between source and target images are established. In the second step, the source image is re‐coloured in a novel optimization framework, which incorporates the extracted content information and the spatial distributions of the target colour styles. A new progressive transfer scheme is proposed to integrate the advantages of both global and local transfer algorithms, as well as avoid the over‐segmentation artefact in the result. Experiments show that with a better understanding of the scene contents, our method well preserves the spatial layout, the colour distribution and the visual coherence in the transfer process. As an interesting extension, our method can also be used to re‐colour video clips with spatially‐varied colour effects.

Color transfer is a practical image editing technology which is useful in various applications. An ideal color transfer algorithm should keep the scene in the source image and apply the color styles of the reference image. All the dominant color styles of the reference image should be presented in the result especially when there are similar contents in the source and reference images.

Level set methods have been extensively used to track the dynamical interfaces between different materials for physically based simulation, geometry modeling, oceanic modeling and other scientific and engineering applications. Due to the inherent Eulerian characteristics, interface evolution based on level set usually suffers from numerical diffusion, sharp feature missing and mass loss. Although some effective methods such as Particle Level Set (PLS) and Marker Level Set (MLS) have been proposed to tackle these difficulties, the complicated correction process and the high computational cost pose severe limitations for real-time applications. In this paper we provide an efficient parallel implementation of the Marker Level Set method on latest graphics hardware. Each step of the MLS method is fully mapped on GPU with an innovative combination of different computation techniques. Relying on GPU's parallelism and flexible programmability, the method provides real-time performance for large size 2D examples and moderate 3D examples, which is significantly faster than previous CPU-based methods.

Natural mountains and valleys are gradually eroded by rainfall and river flows. Physically-based modeling of this complex phenomenon is a major concern in producing realistic synthesized terrains. However, despite some recent improvements, existing algorithms are still computationally expensive, leading to a time-consuming process fairly impractical for terrain designers and 3D artists. In this paper, we present a new method to model the hydraulic erosion phenomenon which runs at interactive rates on today's computers. The method is based on the velocity field of the running water, which is created with an efficient shallow-water fluid model. The velocity field is used to calculate the erosion and deposition process, and the sediment transportation process. The method has been carefully designed to be implemented totally on GPU, and thus takes full advantage of the parallelism of current graphics hardware. Results from experiments demonstrate that the proposed method is effective and efficient. It can create realistic erosion effects by rainfall and river flows, and produce fast simulation results for terrains with large sizes.