T2I-FactualBench: Benchmarking the Factuality of Text-to-Image Models with Knowledge-Intensive Concepts

Evaluating the quality of synthesized images remains a significant challenge in the development of text-to-image (T2I) generation. Most existing studies in this area primarily focus on evaluating text-image alignment, image quality, and object composition capabilities, with comparatively fewer studies addressing the evaluation of the factuality of T2I models, particularly when the concepts involved are knowledge-intensive. To mitigate this gap, we present T2I-FactualBench in this work - the largest benchmark to date in terms of the number of concepts and prompts specifically designed to evaluate the factuality of knowledge-intensive concept generation. T2I-FactualBench consists of a three-tiered knowledge-intensive text-to-image generation framework, ranging from the basic memorization of individual knowledge concepts to the more complex composition of multiple knowledge concepts. We further introduce a multi-round visual question answering (VQA) based evaluation framework to assess the factuality of three-tiered knowledge-intensive text-to-image generation tasks. Experiments on T2I-FactualBench indicate that current state-of-the-art (SOTA) T2I models still leave significant room for improvement.

Multi-Modal Intelligent Channel Modeling: A New Modeling Paradigm via Synesthesia of Machines

In the future sixth-generation (6G) era, to support accurate localization sensing and efficient communication link establishment for intelligent agents, a comprehensive understanding of the surrounding environment and proper channel modeling are indispensable. The existing method, which solely exploits radio frequency (RF) communication information, is difficult to accomplish accurate channel modeling. Fortunately, multi-modal devices are deployed on intelligent agents to obtain environmental features, which could further assist in channel modeling. Currently, some research efforts have been devoted to utilizing multi-modal information to facilitate channel modeling, while still lack a comprehensive review. To fill this gap, we embark on an initial endeavor with the goal of reviewing multi-modal intelligent channel modeling (MMICM) via Synesthesia of Machines (SoM). Compared to channel modeling approaches that solely utilize RF communication information, the utilization of multi-modal information can provide a more in-depth understanding of the propagation environment around the transceiver, thus facilitating more accurate channel modeling. First, this paper introduces existing channel modeling approaches from the perspective of the channel modeling evolution. Then, we have elaborated and investigated recent advances in the topic of capturing typical channel characteristics and features, i.e., channel non-stationarity and consistency, by characterizing the mathematical, spatial, coupling, and mapping relationships. In addition, applications that can be supported by MMICM are summarized and analyzed. To corroborate the superiority of MMICM via SoM, we give the simulation result and analysis. Finally, some open issues and potential directions for the MMICM are outlined from the perspectives of measurements, modeling, and applications.

MARS: Mixture of Auto-Regressive Models for Fine-grained Text-to-image Synthesis

Auto-regressive models have made significant progress in the realm of language generation, yet they do not perform on par with diffusion models in the domain of image synthesis. In this work, we introduce MARS, a novel framework for T2I generation that incorporates a specially designed Semantic Vision-Language Integration Expert (SemVIE). This innovative component integrates pre-trained LLMs by independently processing linguistic and visual information, freezing the textual component while fine-tuning the visual component. This methodology preserves the NLP capabilities of LLMs while imbuing them with exceptional visual understanding. Building upon the powerful base of the pre-trained Qwen-7B, MARS stands out with its bilingual generative capabilities corresponding to both English and Chinese language prompts and the capacity for joint image and text generation. The flexibility of this framework lends itself to migration towards any-to-any task adaptability. Furthermore, MARS employs a multi-stage training strategy that first establishes robust image-text alignment through complementary bidirectional tasks and subsequently concentrates on refining the T2I generation process, significantly augmenting text-image synchrony and the granularity of image details. Notably, MARS requires only 9% of the GPU days needed by SD1.5, yet it achieves remarkable results across a variety of benchmarks, illustrating the training efficiency and the potential for swift deployment in various applications.

Scatterer Recognition from LiDAR Point Clouds for Environment-Embedded Vehicular Channel Modeling via Synesthesia of Machines

In this paper, a novel environment-embedded vehicular channel model is proposed by scatterer recognition from light detection and ranging (LiDAR) point clouds via Synesthesia of Machines (SoM). To provide a robust data foundation, a new intelligent sensing-communication integration dataset in vehicular urban scenarios is constructed. Based on the constructed dataset, the complex SoM mechanism, i.e., mapping relationship between scatterers in electromagnetic space and LiDAR point clouds in physical environment, is explored via multilayer perceptron (MLP) with electromagnetic propagation mechanism. By using LiDAR point clouds to implement scatterer recognition, channel non-stationarity and consistency are modeled in an environment-embedded manner. Using ray-tracing (RT)-based results as the ground truth, the scatterer recognition accuracy exceeds 90%. The accuracy of the proposed model is further verified by the close fit between simulation results and RT results.

Detecting and Mitigating Hallucination in Large Vision Language Models via Fine-Grained AI Feedback

The rapidly developing Large Vision Language Models (LVLMs) have shown notable capabilities on a range of multi-modal tasks, but still face the hallucination phenomena where the generated texts do not align with the given contexts, significantly restricting the usages of LVLMs. Most previous work detects and mitigates hallucination at the coarse-grained level or requires expensive annotation (e.g., labeling by proprietary models or human experts). To address these issues, we propose detecting and mitigating hallucinations in LVLMs via fine-grained AI feedback. The basic idea is that we generate a small-size sentence-level hallucination annotation dataset by proprietary models, whereby we train a hallucination detection model which can perform sentence-level hallucination detection, covering primary hallucination types (i.e., object, attribute, and relationship). Then, we propose a detect-then-rewrite pipeline to automatically construct preference dataset for training hallucination mitigating model. Furthermore, we propose differentiating the severity of hallucinations, and introducing a Hallucination Severity-Aware Direct Preference Optimization (HSA-DPO) for mitigating hallucination in LVLMs by incorporating the severity of hallucinations into preference learning. Extensive experiments demonstrate the effectiveness of our method.

A LiDAR-Aided Channel Model for Vehicular Intelligent Sensing-Communication Integration

In this paper, a novel channel modeling approach, named light detection and ranging (LiDAR)-aided geometry-based stochastic modeling (LA-GBSM), is developed. Based on the developed LA-GBSM approach, a new millimeter wave (mmWave) channel model for sixth-generation (6G) vehicular intelligent sensing-communication integration is proposed, which can support the design of intelligent transportation systems (ITSs). The proposed LA-GBSM is accurately parameterized under high, medium, and low vehicular traffic density (VTD) conditions via a sensing-communication simulation dataset with LiDAR point clouds and scatterer information for the first time. Specifically, by detecting dynamic vehicles and static building/tress through LiDAR point clouds via machine learning, scatterers are divided into static and dynamic scatterers. Furthermore, statistical distributions of parameters, e.g., distance, angle, number, and power, related to static and dynamic scatterers are quantified under high, medium, and low VTD conditions. To mimic channel non-stationarity and consistency, based on the quantified statistical distributions, a new visibility region (VR)-based algorithm in consideration of newly generated static/dynamic scatterers is developed. Key channel statistics are derived and simulated. By comparing simulation results and ray-tracing (RT)-based results, the utility of the proposed LA-GBSM is verified.

Generating Realistic Counterfactuals for Retinal Fundus and OCT Images using Diffusion Models

Counterfactual reasoning is often used in clinical settings to explain decisions or weigh alternatives. Therefore, for imaging based specialties such as ophthalmology, it would be beneficial to be able to create counterfactual images, illustrating answers to questions like "If the subject had had diabetic retinopathy, how would the fundus image have looked?". Here, we demonstrate that using a diffusion model in combination with an adversarially robust classifier trained on retinal disease classification tasks enables the generation of highly realistic counterfactuals of retinal fundus images and optical coherence tomography (OCT) B-scans. The key to the realism of counterfactuals is that these classifiers encode salient features indicative for each disease class and can steer the diffusion model to depict disease signs or remove disease-related lesions in a realistic way. In a user study, domain experts also found the counterfactuals generated using our method significantly more realistic than counterfactuals generated from a previous method, and even indistinguishable from real images.

Intelligent Multi-Modal Sensing-Communication Integration: Synesthesia of Machines

In the era of sixth-generation (6G) wireless communications, integrated sensing and communications (ISAC) is recognized as a promising solution to upgrading the physical system by endowing wireless communications with sensing capability. Existing ISAC is mainly oriented to static scenarios with radio-frequency sensors being the primary participants, thus lacking a comprehensive environment feature characterization and facing a severe performance bottleneck in dynamic environments. In light of this, we generalize the concept of ISAC by mimicking human synesthesia to support intelligent multi-modal sensing-communication integration. The so-termed Synesthesia of Machines (SoM) is not only oriented to generic scenarios, but also particularly suitable for solving challenges arising from dynamic scenarios. We commence by justifying the necessity and potentials of SoM. Subsequently, we offer the definition of SoM and zoom into the specific operating modes, followed by discussions of the state-of-the-art, corresponding objectives, and challenges. To facilitate SoM research, we overview the prerequisite of SoM research, that is, mixed multi-modal (MMM) datasets, and introduce our work. Built upon the MMM datasets, we introduce the mapping relationships between multi-modal sensing and communications, and discuss how channel modeling can be customized to support the exploration of such relationships. Afterwards, we delve into the current research state and implementing challenges of SoM-enhance-based and SoM-concert-based applications. We first overview the SoM-enhance-based communication system designs and present simulation results related to dual-function waveform and predictive beamforming design. Afterwards, we review the recent advances of SoM-concert for single-agent and multi-agent environment sensing. Finally, we propose some open issues and potential directions.

M$^3$SC: A Generic Dataset for Mixed Multi-Modal Sensing and Communication Integration

The sixth generation (6G) of mobile communication system is witnessing a new paradigm shift, i.e., integrated sensing-communication system. A comprehensive dataset is a prerequisite for 6G integrated sensing-communication research. This paper develops a novel simulation dataset, named M3SC, for mixed multi-modal (MMM) sensing-communication integration, and the generation framework of the M3SC dataset is further given. To obtain multi-modal sensory data in physical space and communication data in electromagnetic space, we utilize AirSim and WaveFarer to collect multi-modal sensory data and exploit Wireless InSite to collect communication data. Furthermore, the in-depth integration and precise alignment of AirSim, WaveFarer, and Wireless InSite are achieved. The M3SC dataset covers various weather conditions, various frequency bands, and different times of the day. Currently, the M3SC dataset contains 1500 snapshots, including 80 RGB images, 160 depth maps, 80 LiDAR point clouds, 256 sets of mmWave waveforms with 8 radar point clouds, and 72 channel impulse response (CIR) matrices per snapshot, thus totaling 120,000 RGB images, 240,000 depth maps, 120,000 LiDAR point clouds, 384,000 sets of mmWave waveforms with 12,000 radar point clouds, and 108,000 CIR matrices. The data processing result presents the multi-modal sensory information and communication channel statistical properties. Finally, the MMM sensing-communication application, which can be supported by the M3SC dataset, is discussed.

DcnnGrasp: Towards Accurate Grasp Pattern Recognition with Adaptive Regularizer Learning

The task of grasp pattern recognition aims to derive the applicable grasp types of an object according to the visual information. Current state-of-the-art methods ignore category information of objects which is crucial for grasp pattern recognition. This paper presents a novel dual-branch convolutional neural network (DcnnGrasp) to achieve joint learning of object category classification and grasp pattern recognition. DcnnGrasp takes object category classification as an auxiliary task to improve the effectiveness of grasp pattern recognition. Meanwhile, a new loss function called joint cross-entropy with an adaptive regularizer is derived through maximizing a posterior, which significantly improves the model performance. Besides, based on the new loss function, a training strategy is proposed to maximize the collaborative learning of the two tasks. The experiment was performed on five household objects datasets including the RGB-D Object dataset, Hit-GPRec dataset, Amsterdam library of object images (ALOI), Columbia University Image Library (COIL-100), and MeganePro dataset 1. The experimental results demonstrated that the proposed method can achieve competitive performance on grasp pattern recognition with several state-of-the-art methods. Specifically, our method even outperformed the second-best one by nearly 15% in terms of global accuracy for the case of testing a novel object on the RGB-D Object dataset.