Achieving flexible and high-fidelity identity-preserved image generation remains formidable, particularly with advanced Diffusion Transformers (DiTs) like FLUX. We introduce InfiniteYou (InfU), one of the earliest robust frameworks leveraging DiTs for this task. InfU addresses significant issues of existing methods, such as insufficient identity similarity, poor text-image alignment, and low generation quality and aesthetics. Central to InfU is InfuseNet, a component that injects identity features into the DiT base model via residual connections, enhancing identity similarity while maintaining generation capabilities. A multi-stage training strategy, including pretraining and supervised fine-tuning (SFT) with synthetic single-person-multiple-sample (SPMS) data, further improves text-image alignment, ameliorates image quality, and alleviates face copy-pasting. Extensive experiments demonstrate that InfU achieves state-of-the-art performance, surpassing existing baselines. In addition, the plug-and-play design of InfU ensures compatibility with various existing methods, offering a valuable contribution to the broader community.

Event-based image reconstruction has achieved remarkable progress, benefiting from the high temporal resolution and high dynamic range of event cameras. However, most event-based methods focus on enhancing sRGB image quality, neglecting the potential of leveraging event data for RAW-to-sRGB conversion. Due to the limitations of camera sensors, images processed through standard ISP pipelines often suffer from motion blur and color distortion in dynamic scenes. In contrast, RAW images preserve uncompressed scene information, integrating event signals at this stage enables finer texture recovery and more accurate color correction. To tackle these challenges, we propose EvRAW, a novel event-assisted RAW-to-sRGB image reconstruction network that integrates event signals to promote high-fidelity sRGB image reconstruction. Specifically, we introduce a Motion-guided Structural Enhancement (MSE) module that extracts motion patterns from event streams and aggregates dynamic features to restore fine textures. Additionally, we propose an Adaptive Color Correction (ACC) module that performs region-wise gamma correction and channel-wise color decoding to enhance color fidelity under complex lighting conditions. To evaluate performance in challenging real-world scenarios, we collect a pixel-aligned RAW-Event dataset specifically for this task. Extensive experiments demonstrate that EvRAW achieves state-of-the-art performance in RAW-to-sRGB reconstruction on both synthetic and real-world datasets.

Large language models (LLMs) have shown remarkable performance across diverse reasoning and generation tasks, and are increasingly deployed as agents in dynamic environments such as code generation and recommendation systems. However, many real-world applications, such as high-frequency trading and real-time competitive gaming, require decisions under strict latency constraints, where faster responses directly translate into higher rewards. Despite the importance of this latency–quality trade-off, it remains underexplored in the context of LLM-based agents. In this work, we present the first systematic study of this trade-off in real-time decision-making tasks. To support our investigation, we introduce two new benchmarks: HFTBench, a high-frequency trading simulation, and StreetFighter, a competitive gaming platform. Our analysis reveals that optimal latency–quality balance varies by task, and that sacrificing quality for lower latency can significantly enhance downstream performance. To address this, we propose FPX, an adaptive framework that dynamically selects model size and quantization level based on real-time demands. Our method achieves the best performance on both benchmarks, improving win rate by up to 80% in Street Fighter and boosting daily yield by up to 26.52% in trading, underscoring the need for latency-aware evaluation and deployment strategies for LLM-based agents. These results demonstrate the critical importance of latency-aware evaluation and deployment strategies for real-world LLM-based agents.

We present a deep learning model, dubbed Glissando-Net, to simultaneously estimate the pose and reconstruct the 3D shape of objects at the category level from a single RGB image. Previous works predominantly focused on either estimating poses (often at the instance level), or reconstructing shapes, but not both. Glissando-Net is composed of two auto-encoders that are jointly trained, one for RGB images and the other for point clouds. We embrace two key design choices in Glissando-Net to achieve a more accurate prediction of the 3D shape and pose of the object given a single RGB image as input. First, we augment the feature maps of the point cloud encoder and decoder with transformed feature maps from the image decoder, enabling effective 2D-3D interaction in both training and prediction. Second, we predict both the 3D shape and pose of the object in the decoder stage. This way, we better utilize the information in the 3D point clouds presented only in the training stage to train the network for more accurate prediction. We jointly train the two encoder-decoders for RGB and point cloud data to learn how to pass latent features to the point cloud decoder during inference. In testing, the encoder of the 3D point cloud is discarded. The design of Glissando-Net is inspired by codeSLAM. Unlike codeSLAM, which targets 3D reconstruction of scenes, we focus on pose estimation and shape reconstruction of objects, and directly predict the object pose and a pose invariant 3D reconstruction without the need of the code optimization step. Extensive experiments, involving both ablation studies and comparison with competing methods, demonstrate the efficacy of our proposed method, and compare favorably with the state-of-the-art.

We have witnessed significant progress in deep learning-based 3D vision, ranging from neural radiance field (NeRF) based 3D representation learning to applications in novel view synthesis (NVS). However, existing scene-level datasets for deep learning-based 3D vision, limited to ei-ther synthetic environments or a narrow selection of real-world scenes, are quite insufficient. This insufficiency not only hinders a comprehensive benchmark of existing methods but also caps what could be explored in deep learning-based 3D analysis. To address this critical gap, we present DL3DV-10K, a large-scale scene dataset, featuring 51.2 million frames from 10,510 videos captured from 65 types of point- of-interest (POI) locations, covering both bounded and unbounded scenes, with different levels of reflection, transparency, and lighting. We conducted a comprehensive benchmark of recent NVS methods on DL3DV-10K, which revealed valuable insights for future research in NVS. In addition, we have obtained encouraging results in a pilot study to learn generalizable NeRF from DL3DV-10K, which manifests the necessity of a large-scale scene-level dataset to forge a path toward a foundation model for learning 3D representation. Our DL3DV-10K dataset, benchmark results, and models will be publicly accessible.

A fundamental problem in learning 3D shapes generative models is that when the generative model is simply fitted to the training data, the resulting synthetic 3D models can present various artifacts. Many of these artifacts are topological in nature, e.g., broken legs, unrealistic thin structures, and small holes. In this paper, we introduce a principled approach that utilizes topological regularization losses on an implicit shape generator to rectify topological artifacts. The objectives are two-fold. The first is to align the persistent diagram (PD) distribution of the training shapes with that of synthetic shapes. The second ensures that the PDs are smooth among adjacent synthetic shapes. We show how to achieve these two objectives using two simple but effective formulations. Specifically, distribution alignment is achieved to learn a generative model of PDs and align this generator with PDs of synthetic shapes. We show how to handle discrete and continuous variabilities of PDs by using a shape-regularization term when performing PD alignment. Moreover, we enforce the smoothness of the PDs using a smoothness loss on the PD generator, which further improves the behavior of PD distribution alignment. Experimental results on ShapeNet show that our approach leads to much better generalization behavior than state-of-the-art implicit shape generators.

Camera calibration is very important for some vision tasks, like rendering 3D scenes, environment reconstruction, and self-localization, etc. In this paper, we propose a framework of multi-scale transformers for camera calibration. With the input of a single image, the multi-scale features output from the model’s backbone are utilized to estimate camera parameters. At the same time, we show that the way of coarse-to-fine is effective to locate global structures and detailed features in the image, by studying the attention response of horizon line estimation. Moreover, deep supervision is proven to get more precise results and accelerated training. Our method outperforms all the state-of-the-art methods by objective and subjective experiments on Google Street View dataset and Pano360.

Abstract The RSA (Rivest–Shamir–Adleman) cryptosystem is a widely used public-key cryptographic algorithm in information systems and computer applications. With the advancement of lattice theory, a technique known as the lattice-based method has emerged as a significant threat to RSA and its variants. This paper aims to conduct a bibliometric analysis of publications in the field of lattice-based attacks on RSA-type cryptosystems. The analysis is based on a dataset of relevant publications retrieved from Scopus and utilizes visualization tools such as CiteSpace and VOSviewer for a thorough overview. In order to understand the research developmental trajectory, we analyze the publication trends over the years, explore cooperation networks at various levels, including country/region, institution, and author, and assess the extent of collaboration, contribution, and productivity within the field. Additionally, author co-citation network and reference co-citation clustering are utilized to enable the identification of significant research achievements, cutting-edge developments, and structural framework. Furthermore, we conduct detailed analysis from a keyword perspective to identify research hotspots and emerging trends. The findings contribute to the existing body of knowledge on lattice-based cryptanalysis of RSA-type cryptosystems. Moreover, this bibliometric analysis serves as a valuable resource for identifying fruitful areas for further exploration and guides future research efforts.

This paper introduces a novel approach for identifying the possible large language models (LLMs) involved in text generation.Instead of adding an additional classification layer to a base LM, we reframe the classification task as a next-token prediction task and directly fine-tune the base LM to perform it.We utilize the Text-to-Text Transfer Transformer (T5) model as the backbone for our experiments.We compared our approach to the more direct approach of utilizing hidden states for classification.Evaluation shows the exceptional performance of our method in the text classification task, highlighting its simplicity and efficiency.Furthermore, interpretability studies on the features extracted by our model reveal its ability to differentiate distinctive writing styles among various LLMs even in the absence of an explicit classifier.We also collected a dataset named OpenLLMText, containing approximately 340k text samples from human and LLMs, including GPT3.

This paper advocates the use of implicit surface representation in autoencoder-based self-supervised 3D representation learning. The most popular and accessible 3D representation, i.e., point clouds, involves discrete samples of the underlying continuous 3D surface. This discretization process introduces sampling variations on the 3D shape, making it challenging to develop transferable knowledge of the true 3D geometry. In the standard autoencoding paradigm, the encoder is compelled to encode not only the 3D geometry but also information on the specific discrete sampling of the 3D shape into the latent code. This is because the point cloud reconstructed by the decoder is considered unacceptable unless there is a perfect mapping between the original and the reconstructed point clouds. This paper introduces the Implicit AutoEncoder (IAE), a simple yet effective method that addresses the sampling variation issue by replacing the commonly-used point-cloud decoder with an implicit decoder. The implicit decoder reconstructs a continuous representation of the 3D shape, independent of the imperfections in the discrete samples. Extensive experiments demonstrate that the proposed IAE achieves state-of-the-art performance across various self-supervised learning benchmarks. Our code is available at https://github.com/SimingYan/IAE.

To improve the efficiency and sustainability of learning deep models, we propose CREST, the first scalable framework with rigorous theoretical guarantees to identify the most valuable examples for training non-convex models, particularly deep networks. To guarantee convergence to a stationary point of a non-convex function, CREST models the non-convex loss as a series of quadratic functions and extracts a coreset for each quadratic sub-region. In addition, to ensure faster convergence of stochastic gradient methods such as (mini-batch) SGD, CREST iteratively extracts multiple mini-batch coresets from larger random subsets of training data, to ensure nearly-unbiased gradients with small variances. Finally, to further improve scalability and efficiency, CREST identifies and excludes the examples that are learned from the coreset selection pipeline. Our extensive experiments on several deep networks trained on vision and NLP datasets, including CIFAR-10, CIFAR-100, TinyImageNet, and SNLI, confirm that CREST speeds up training deep networks on very large datasets, by 1.7x to 2.5x with minimum loss in the performance. By analyzing the learning difficulty of the subsets selected by CREST, we show that deep models benefit the most by learning from subsets of increasing difficulty levels.

Terrains are visually prominent and commonly needed objects in many computer graphics applications. While there are many algorithms for synthetic terrain generation, it is rather difficult to assess the realism of a generated output. This article presents a first step toward the direction of perceptual evaluation for terrain models. We gathered and categorized several classes of real terrains, and we generated synthetic terrain models using computer graphics methods. The terrain geometries were rendered by using the same texturing, lighting, and camera position. Two studies on these image sets were conducted, ranking the terrains perceptually, and showing that the synthetic terrains are perceived as lacking realism compared to the real ones. We provide insight into the features that affect the perceived realism by a quantitative evaluation based on localized geomorphology-based landform features (geomorphons) that categorize terrain structures such as valleys, ridges, hollows, and so forth. We show that the presence or absence of certain features has a significant perceptual effect. The importance and presence of the terrain features were confirmed by using a generative deep neural network that transferred the features between the geometric models of the real terrains and the synthetic ones. The feature transfer was followed by another perceptual experiment that further showed their importance and effect on perceived realism. We then introduce Perceived Terrain Realism Metrics (PTRM), which estimates human-perceived realism of a terrain represented as a digital elevation map by relating the distribution of terrain features with their perceived realism. This metric can be used on a synthetic terrain, and it will output an estimated level of perceived realism. We validated the proposed metrics on real and synthetic data and compared them to the perceptual studies.

The problem of long-tailed recognition, where the number of examples per class is highly unbalanced, is considered. It is hypothesized that the well known tendency of standard classifier training to overfit to popular classes can be exploited for effective transfer learning. Rather than eliminating this overfitting, e.g. by adopting popular class-balanced sampling methods, the learning algorithm should instead leverage this overfitting to transfer geometric information from popular to low-shot classes. A new classifier architecture, GistNet, is proposed to support this goal, using constellations of classifier parameters to encode the class geometry. A new learning algorithm is then proposed for GeometrIc Structure Transfer (GIST), with resort to a combination of loss functions that combine class-balanced and random sampling to guarantee that, while overfitting to the popular classes is restricted to geometric parameters, it is leveraged to transfer class geometry from popular to few-shot classes. This enables better generalization for few-shot classes without the need for the manual specification of class weights, or even the explicit grouping of classes into different types. Experiments on two popular long-tailed recognition datasets show that GistNet outperforms existing solutions to this problem.

The problem of open-set recognition is considered. While previous approaches only consider this problem in the context of large-scale classifier training, we seek a unified solution for this and the low-shot classification setting. It is argued that the classic softmax classifier is a poor solution for open-set recognition, since it tends to overfit on the training classes. Randomization is then proposed as a solution to this problem. This suggests the use of meta-learning techniques, commonly used for few-shot classification, for the solution of open-set recognition. A new oPen sEt mEta LEaRning (PEELER) algorithm is then introduced. This combines the random selection of a set of novel classes per episode, a loss that maximizes the posterior entropy for examples of those classes, and a new metric learning formulation based on the Mahalanobis distance. Experimental results show that PEELER achieves state of the art open set recognition performance for both few-shot and large-scale recognition. On CIFAR and miniImageNet, it achieves substantial gains in seen/unseen class detection AUROC for a given seen-class classification accuracy.

Interactions play a key role in understanding objects and scenes for both virtual and real-world agents. We introduce a new general representation for proximal interactions among physical objects that is agnostic to the type of objects or interaction involved. The representation is based on tracking particles on one of the participating objects and then observing them with sensors appropriately placed in the interaction volume or on the interaction surfaces. We show how to factorize these interaction descriptors and project them into a particular participating object so as to obtain a new functional descriptor for that object, its interaction landscape , capturing its observed use in a spatiotemporal framework. Interaction landscapes are independent of the particular interaction and capture subtle dynamic effects in how objects move and behave when in functional use. Our method relates objects based on their function, establishes correspondences between shapes based on functional key points and regions, and retrieves peer and partner objects with respect to an interaction.

Tumors are developed and worsen with the accumulated mutations on DNA sequences during tumorigenesis. Identifying the temporal order of gene mutations in cancer initiation and development is a challenging topic. It not only provides a new insight into the study of tumorigenesis at the level of genome sequences but also is an effective tool for early diagnosis of tumors and preventive medicine. In this paper, we develop a novel method to accurately estimate the sequential order of gene mutations during tumorigenesis from genome sequencing data based on Markov chain model as TOMC (Temporal Order based on Markov Chain), and also provide a new criterion to further infer the order of samples or patients, which can characterize the severity or stage of the disease. We applied our method to the analysis of tumors based on several high-throughput datasets. Specifically, first, we revealed that tumor suppressor genes (TSG) tend to be mutated ahead of oncogenes, which are considered as important events for key functional loss and gain during tumorigenesis. Second, the comparisons of various methods demonstrated that our approach has clear advantages over the existing methods due to the consideration on the effect of mutation dependence among genes, such as co-mutation. Third and most important, our method is able to deduce the ordinal sequence of patients or samples to quantitatively characterize their severity of tumors. Therefore, our work provides a new way to quantitatively understand the development and progression of tumorigenesis based on high throughput sequencing data.