Manipulation of facial images to meet specific controls such as pose, expression, and lighting, also known as face rigging, is a complex task in computer vision. Existing methods are limited by their reliance on image datasets, which necessitates individual-specific fine-tuning and limits their ability to retain fine-grained identity and semantic details, reducing practical usability. To overcome these limitations, we introduce ControlFace, a novel face rigging method conditioned on 3DMM renderings that enables flexible, high-fidelity control. We employ a dual-branch U-Nets: one, referred to as FaceNet, captures identity and fine details, while the other focuses on generation. To enhance control precision, the control mixer module encodes the correlated features between the target-aligned control and reference-aligned control, and a novel guidance method, reference control guidance, steers the generation process for better control adherence. By training on a facial video dataset, we fully utilize FaceNet’s rich representations while ensuring control adherence. Extensive experiments demonstrate ControlFace’s superior performance in identity preservation and control precision, highlighting its practicality. Please see the project website: https://cvlab-kaist.github.io/ControlFace/.

JoonHo Lee, JuYoun Son, Juree Seok, Wooseok Jang, Yeong-Dae Kwon. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Recent guidance methods in diffusion models steer reverse sampling by perturbing the model to construct an implicit weak model and guide generation away from it. Among these approaches, attention perturbation has demonstrated strong empirical performance in unconditional scenarios where classifier-free guidance is not applicable. However, existing attention perturbation methods lack principled approaches for determining where perturbations should be applied, particularly in Diffusion Transformer (DiT) architectures where quality-relevant computations are distributed across layers. In this paper, we investigate the granularity of attention perturbations, ranging from the layer level down to individual attention heads, and discover that specific heads govern distinct visual concepts such as structure, style, and texture quality. Building on this insight, we propose ``HeadHunter", a systematic framework for iteratively selecting attention heads that align with user-centric objectives, enabling fine-grained control over generation quality and visual attributes. In addition, we introduce SoftPAG, which linearly interpolates each selected head’s attention map toward an identity matrix, providing a continuous knob to tune perturbation strength and suppress artifacts. Our approach not only mitigates the oversmoothing issues of existing layer-level perturbation but also enables targeted manipulation of specific visual styles through compositional head selection. We validate our method on modern large-scale DiT-based text-to-image models including Stable Diffusion 3 and FLUX.1, demonstrating superior performance in both general quality enhancement and style-specific guidance. Our work provides the first head-level analysis of attention perturbation in diffusion models, uncovering interpretable specialization within attention layers and enabling practical design of effective perturbation strategies.

Learning robust vision models that perform well in out-of-distribution (OOD) situations is an important task for model deployment in real-world settings. Despite extensive research in this field, many proposed methods have only shown minor performance improvements compared to the simplest empirical risk minimization (ERM) approach, which was evaluated on a benchmark with a limited hy-perparameter search space. Our focus in this study is on leveraging the knowledge of large pretrained models to improve handling of OOD scenarios and tackle domain generalization problems. However, prior research has revealed that naively fine-tuning a large pretrained model can impair OOD robustness. Thus, we employ parameter-efficient fine-tuning (PEFT) techniques to effectively preserve OOD robustness while working with large models. Our extensive experiments and analysis confirm that the most effective approaches involve ensembling diverse models and increasing the scale of pretraining. As a result, we achieve state-of-the-art performance in domain generalization tasks. Our code and project page are available at: https://cvlab-kaist.github.io/MoA

Text-to-3D generation has shown rapid progress in recent days with the advent of score distillation sampling (SDS), a methodology of using pretrained text-to-2D diffusion models to optimize a neural radiance field (NeRF) in a zero-shot setting. However, the lack of 3D awareness in the 2D diffusion model often destabilizes previous methods from generating a plausible 3D scene. To address this issue, we propose 3DFuse, a novel framework that incorporates 3D awareness into the pretrained 2D diffusion model, enhancing the robustness and 3D consistency of score distillation-based methods. Specifically, we introduce a consistency injection module which constructs a 3D point cloud from the text prompt and utilizes its projected depth map at given view as a condition for the diffusion model. The 2D diffusion model, through its generative capability, robustly infers dense structure from the sparse point cloud depth map and generates a geometrically consistent and coherent 3D scene. We also introduce a new technique called semantic coding that reduces semantic ambiguity of the text prompt for improved results. Our method can be easily adapted to various text-to-3D baselines, and we experimentally demonstrate how our method notably enhances the 3D consistency of generated scenes in comparison to previous baselines, achieving state-of-the-art performance in geometric robustness and fidelity.

Assessing response quality to instructions in language models is vital but challenging due to the complexity of human language across different contexts. This complexity often results in ambiguous or inconsistent interpretations, making accurate assessment difficult. To address this issue, we propose a novel Uncertainty-aware Reward Model (URM) that introduces a robust uncertainty estimation for the quality of paired responses based on Bayesian approximation. Trained with preference datasets, our uncertainty-enabled proxy not only scores rewards for responses but also evaluates their inherent uncertainty. Empirical results demonstrate significant benefits of incorporating the proposed proxy into language model training. Our method boosts the instruction following capability of language models by refining data curation for training and improving policy optimization objectives, thereby surpassing existing methods by a large margin on benchmarks such as Vicuna and MT-bench. These findings highlight that our proposed approach substantially advances language model training and paves a new way of harnessing uncertainty within language models.

Text-to-3D generation has achieved significant success by incorporating powerful 2D diffusion models, but insufficient 3D prior knowledge also leads to the inconsistency of 3D geometry. Recently, since large-scale multi-view datasets have been released, fine-tuning the diffusion model on the multi-view datasets becomes a mainstream to solve the 3D inconsistency problem. However, it has confronted with fundamental difficulties regarding the limited quality and diversity of 3D data, compared with 2D data. To sidestep these trade-offs, we explore a retrieval-augmented approach tailored for score distillation, dubbed ReDream. We postulate that both expressiveness of 2D diffusion models and geometric consistency of 3D assets can be fully leveraged by employing the semantically relevant assets directly within the optimization process. To this end, we introduce novel framework for retrieval-based quality enhancement in text-to-3D generation. We leverage the retrieved asset to incorporate its geometric prior in the variational objective and adapt the diffusion model's 2D prior toward view consistency, achieving drastic improvements in both geometry and fidelity of generated scenes. We conduct extensive experiments to demonstrate that ReDream exhibits superior quality with increased geometric consistency. Project page is available at https://ku-cvlab.github.io/ReDream/.

Denoising diffusion models (DDMs) have attracted attention for their exceptional generation quality and diversity. This success is largely attributed to the use of class- or text-conditional diffusion guidance methods, such as classifier and classifier-free guidance. In this paper, we present a more comprehensive perspective that goes beyond the traditional guidance methods. From this generalized perspective, we introduce novel condition- and training-free strategies to enhance the quality of generated images. As a simple solution, blur guidance improves the suitability of intermediate samples for their fine-scale information and structures, enabling diffusion models to generate higher quality samples with a moderate guidance scale. Improving upon this, Self-Attention Guidance (SAG) uses the intermediate self-attention maps of diffusion models to enhance their stability and efficacy. Specifically, SAG adversarially blurs only the regions that diffusion models attend to at each iteration and guides them accordingly. Our experimental re sults show that our SAG improves the performance of various diffusion models, including ADM, IDDPM, Stable Diffusion, and DiT. Moreover, combining SAG with conventional guidance methods leads to further improvement.

This paper presents a dexterous three degree-of-freedom (3-DOF) wrist mechanism with a large range of motion and uniform manipulability without singular points throughout the entire range of motion. It has a 2-DOF spherical pure rolling joint surrounded by two pairs of actuating wires, the motions of which directly represent the quaternion values of the joint; this joint is therefore named the quaternion joint. Based on this property, it has simple and clear forward and inverse kinematics and high manipulability. By adding a 1-DOF rotation joint at the distal end of the quaternion joint, it can be extended to a 3-DOF joint mechanism. To precisely approximate the spherical pure rolling motion in a confined central space, a novel parallel mechanism composed of three identical supporting linkages was introduced. Unlike conventional parallel mechanisms, it has a compact and simple structure with no interference among the supporting linkages. Because the wrist mechanism is a tendon-driven mechanism, and is thus suitable for lightweight manipulators, it is mounted to a low-inertia manipulator with high stiffness and strength, namely, LIMS2-AMBIDEX, which is an improved version of the authors' previous research. The basic concept and thorough theoretical analysis of the wrist mechanism are described herein, and the simulations and experiments conducted for a quantitative validation are presented.

In this paper, we propose a stereo matching algorithm with occlusion handling. In order to detect occlusion, we obtain an initial disparity map via optimization based on modified constant-space belief propagation (CSBP). Such a method is advantageous due to its low complexity. The initial disparity maps provide clue for occlusion detection. From such clue, an energy function for occlusion detection is defined and optimized by energy minimization framework. We classify occlusion into two types from the obtained occlusion map and apply suitable occlusion handling process, respectively. The proposed occlusion handling method based on the potential energy function extends disparity values of visible pixels to occluded pixels. Experimental results show that generated disparity map of the proposed method has satisfactory quality.

This paper proposes the hybrid simplex algorithm (SA)-harmony search(HS) Method. HS method is, the evolutionary algorithm, conceptualized using the musical process of searching for optimization problems. SA helps HS find optimization solution more accurately and quickly. In this paper, the performances of proposed algorithm are compared with the original HS method and other algorithms through unconstrained functions and constrained functions.

In this paper, we focus on the optimization of fuzzy clustering. Particle Swarm Optimizations (PSO) is used for optimizing the algorithms. PSO is an algorithm which takes a cue from nature’s bird flock or fish school and is known to have superior ability in search and fast convergence. But it might be difficult to find global optimal solution of the fuzzy clustering when it comes to complex higher dimensions. So we optimize the fuzzy clustering using Predator Prey Particle Swarm Optimizations (PPPSO). The concept of PPPSO is that predators chase the center of prey’s swarm, and preys escape from predators, in order to avoid local optimal solutions and find global optimal solution efficiently. The performance of fuzzy c-means (FCM), particle swarm fuzzy clustering (PSFC) and predator prey particle swarm fuzzy clustering (PPPSFC) are compared. Through experiments, we show that the proposed algorithm has the best performance among them.