论文检索

Yi Yao, Chan-Feng Hsu, Jhe-Hao Lin, Hongxia Xie, Terence Lin, Yi-Ning Huang, Hong-Han Shuai, Wen-Huang Cheng

In spite of recent advancements in text-to-image generation, it still has limitations when it comes to complex, imaginative text prompts. Due to the limited exposure to diverse and complex data in their training sets, text-to-image models often struggle to comprehend the semantics of these difficult prompts, leading to the generation of irrelevant images. This work explores how diffusion models can process and generate images based on prompts requiring artistic creativity or specialized knowledge. Recognizing the absence of a dedicated evaluation framework for such tasks, we introduce a new benchmark, the Realistic-Fantasy Benchmark (RFBench), which blends scenarios from both realistic and fantastical realms. Accordingly, for reality and fantasy scene generation, we propose an innovative training-free approach, Realistic-Fantasy Network (RFNet), that integrates diffusion models with LLMs. Through our proposed RFBench, extensive human evaluations coupled with GPT-based compositional assessments have demonstrated our approach's superiority over other state-of-the-art methods.

Minguk Kang, Richard Zhang, Connelly Barnes, Sylvain Paris, Suha Kwak, Jaesik Park, Eli Shechtman, Jun-Yan Zhu, Taesung Park

We propose a method to distill a complex multistep diffusion model into a single-step student model, dramatically accelerating inference while preserving image quality. Our approach interprets the diffusion distillation as a paired image-to-image translation task, using noise-to-image pairs of the diffusion model’s ODE trajectory. For efficient regression loss computation, we propose E-LatentLPIPS, a Perceptual loss in the latent space of the diffusion model with an ensemble of augmentations. Despite dataset construction costs, E-LatentLPIPS converges more efficiently than many existing distillation methods. Furthermore, we adapt a diffusion model to construct a multi-scale discriminator with a text alignment loss to build an effective conditional GAN-based formulation. We demonstrate that our one-step generator outperforms all published diffusion distillation models on the zero-shot COCO2014 benchmark.

Sumit Sarin, Utkarsh Mall, Purva Tendulkar, Carl Vondrick

Does our behavior change when we speak over video calls? Given two unpaired sets of videos of people, we seek to automatically find spatio-temporal patterns that are distinctive of each set. Existing methods use discriminative approaches and perform post-hoc explainability analysis. Such methods are insufficient as they are unable to provide insights beyond obvious dataset biases, and the explanations are useful only if humans themselves are good at the task. Instead, we tackle the problem through the lens of generative domain translation: our method generates a detailed report of learned, input-dependent spatio-temporal features and the extent to which they vary between the domains. We demonstrate that our method can discover behavioral differences between conversing face-to-face (F2F) and on video-calls (VCs). We also show the applicability of our method on discovering differences in presidential communication styles. Additionally, we are able to predict temporal change-points in videos that decouple expressions in an unsupervised way, and increase the interpretability and usefulness of our model. Finally, our method, being generative, can be used to transform a video call to appear as if it were recorded in a F2F setting. Experiments and visualizations show our approach is able to discover a range of behaviors, taking a step towards deeper understanding of human behaviors.

Bingyu Xin, Meng Ye, Leon Axel, Dimitris N. Metaxas

Magnetic Resonance Imaging (MRI) is a widely used imaging modality for clinical diagnostics and the planning of surgical interventions. Accelerated MRI seeks to mitigate the inherent limitation of long scanning time by reducing the amount of raw $k$-space data required for image reconstruction. Recently, the deep unrolled model (DUM) has demonstrated significant effectiveness and improved interpretability for MRI reconstruction, by truncating and unrolling the conventional iterative reconstruction algorithms with deep neural networks. However, the potential of DUM for MRI reconstruction has not been fully exploited. In this paper, we first enhance the gradient and information flow within and across iteration stages of DUM, then we highlight the importance of using various adjacent information for accurate and memory-efficient sensitivity map estimation and improved multi-coil MRI reconstruction. Extensive experiments on several public MRI reconstruction datasets show that our method outperforms existing MRI reconstruction methods by a large margin. The code will be released publicly.

Saad Manzur, Wayne B Hayes

Existing 2D-to-3D pose lifting networks suffer from poor performance in cross-dataset benchmarks. Although 2D keypoints joined by "stick-figure’’ limbs is the dominant trend, stick-figures do not preserve occlusion information that is inherent in an image, resulting in significant ambiguities that are ruled out when occlusion information is present. In addition, datasets with ground truth 3D poses are much harder to obtain in contrast to similar human annotated 2D datasets. To address these issues, we propose to replace stick figures with abstract images ---figures with opaque limbs that preserve occlusion information while implicitly encoding joint locations. We then break down the pose estimation task into two stages: (1) Generating an abstract image from a real image, and (2) garnering the pose from the abstract image. Crucially, given the GT 3D keypoints for a particular pose, we can synthesize an arbitrary number of abstract images of the same pose as seen from arbitrary cameras, even without a part map. Given a set of 3D GT keypoints, this allows training of Stage (2) on an unlimited dataset without over-training, which in turn allows us to correctly interpret poses from arbitrary viewpoints not included in the original dataset. Additionally, our unlimited training of Stage 2 allows good generalizations across datasets, demonstrated through a significant improvement in cross-dataset benchmarks, while still showing competitive performance in same-dataset benchmark.

Zhaoyang Liu, Zeqiang Lai, Zhangwei Gao, erfei cui, Ziheng Li, Xizhou Zhu, Lewei Lu, Qifeng Chen, Yu Qiao, Jifeng Dai 等

We present ControlLLM, a novel framework that enables large language models (LLMs) to utilize multi-modal tools for solving complex real-world tasks. Despite the remarkable performance of LLMs, they still struggle with tool invocation due to ambiguous user prompts, inaccurate tool selection and parameterization, and inefficient tool scheduling. To overcome these challenges, our framework comprises three key components: (1) a \textit{task decomposer} that breaks down a complex task into clear subtasks with well-defined inputs and outputs; (2) a \textit{Thoughts-on-Graph (ToG) paradigm} that searches the optimal solution path on a pre-built tool graph, which specifies the parameter and dependency relations among different tools; and (3) an \textit{execution engine with a rich toolbox} that interprets the solution path and runs the tools efficiently on different computational devices. We evaluate our framework on diverse tasks involving image, audio, and video processing, demonstrating its superior accuracy, efficiency, and versatility compared to existing methods.

Jinxing Zhou, Dan Guo, Yuxin Mao, Yiran Zhong, Xiaojun Chang, Meng Wang

Audio-Visual Video Parsing (AVVP) task aims to detect and temporally locate events within audio and visual modalities. Multiple events can overlap in the timeline, making identification challenging. While traditional methods usually focus on improving the early audio-visual encoders to embed more effective features, the decoding phase – crucial for final event classification, often receives less attention. We aim to advance the decoding phase and improve its interpretability. Specifically, we introduce a new decoding paradigm, label semantic-based projection (LEAP), that employs labels texts of event categories, each bearing distinct and explicit semantics, for parsing potentially overlapping events. LEAP works by iteratively projecting encoded latent features of audio/visual segments onto semantically independent label embeddings. This process, enriched by modeling cross-modal (audio/visuallabel) interactions, gradually disentangles event semantics within video segments to refine relevant label embeddings, guaranteeing a more discriminative and interpretable decoding process. To facilitate the LEAP paradigm, we propose a semantic-aware optimization strategy, which includes a novel audio-visual semantic similarity loss function. This function leverages the Intersection over Union of audio and visual events (EIoU) as a novel metric to calibrate audio-visual similarities at the feature level, accommodating the varied event densities across modalities. Extensive experiments demonstrate the superiority of our method, achieving new state-of-the-art performance for AVVP and also enhancing the relevant audio-visual event localization task.

Kunyu Peng, Jia Fu, Kailun Yang, Di Wen, Yufan Chen, Ruiping Liu, Junwei Zheng, Jiaming Zhang, Saquib Sarfraz, Rainer Stiefelhagen 等

We introduce a new task called Referring Atomic Video Action Recognition (RAVAR), aimed at identifying atomic actions of a particular person based on a textual description and the video data of this person. This task differs from traditional action recognition and localization, where predictions are delivered for all present individuals. In contrast, we focus on recognizing the correct atomic action of a specific individual, guided by text. To explore this task, we present the RefAVA dataset, containing 36,630 instances with manually annotated textual descriptions of the individuals. To establish a strong initial benchmark, we implement and validate baselines from various domains, e.g., atomic action localization, video question answering, and text-video retrieval. Since these existing methods underperform on RAVAR, we introduce RefAtomNet -- a novel cross-stream attention-driven method specialized for the unique challenges of RAVAR: the need to interpret a textual referring expression for the targeted individual, utilize this reference to guide the spatial localization and harvest the prediction of the atomic actions for the referring person. The key ingredients are: (1) a multi-stream architecture that connects video, text, and a new location-semantic stream, and (2) cross-stream agent attention fusion and agent token fusion which amplify the most relevant information across these streams and consistently surpasses standard attention-based fusion on RAVAR. Extensive experiments demonstrate the effectiveness of RefAtomNet and its building blocks for recognizing the action of the described individual. The dataset and code will be made publicly available.

Yuhuan Yang, Chaofan Ma, Jiangchao Yao, Zhun Zhong, Ya Zhang, Yanfeng Wang

Referring Image Segmentation (RIS) leveraging transformers has achieved great success on the interpretation of complex visual-language tasks. However, the quadratic computation cost makes it difficult in capturing long-range visual-language dependencies, which is particularly important for the context of large-size images with long textual descriptions. Fortunately, Mamba addresses this with efficient linear complexity in processing. However, directly applying Mamba to multi-modal interactions presents challenges, primarily due to inadequate channel interactions for the effective fusion of multi-modal data. In this paper, we propose \methodname, a novel RIS architecture that integrates the efficiency of Mamba with a multi-modal Mamba Twister block. The Mamba Twister explicitly models image-text interaction, and fuses textual and visual features through its unique channel and spatial twisting mechanism. We achieve state-of-the-art on all three benchmarks. Moreover, we conduct thorough analyses of \methodname and discuss other fusion designs using Mamba. These provide valuable perspectives for future research. The code will be released upon publication.

Zhonghan Zhao, Xuan Wang, Wenhao Chai, Boyi Li, Shengyu Hao, Shidong Cao, Tian Ye, Gaoang Wang

Large language models (LLMs) have achieved impresx0002sive progress on several open-world tasks. Recently, usx0002ing LLMs to build embodied agents has been a hotspot. In this paper, we propose STEVE, a comprehensive and visionx0002ary embodied agent in the Minecraft virtual environment. STEVE consists of three key components: vision perception, language instruction, and code action. Vision perception involves the interpretation of visual information in the envix0002ronment, which is then integrated into the LLMs component with agent state and task instruction. Language instrucx0002tion is responsible for iterative reasoning and decomposx0002ing complex tasks into manageable guidelines. Code action generates executable skill actions based on retrieval in skill database, enabling the agent to interact effectively within the Minecraft environment. We also collect STEVE-21K dataset, which includes 600+ vision-environment pairs, 20K knowledge question-answering pairs, and 200+ skillx0002code pairs. We conduct continuous block search, knowledge question and answering, and tech tree mastery to evaluate the performance. Extensive experiments show that STEVE achieves at most 1.5× faster unlocking key tech trees and 2.5× quicker in block search tasks compared to previous state-of-the-art methods.

Yiming Huang, WEILIN WAN, Yue Yang, Chris Callison-Burch, Mark Yatskar, Lingjie Liu

Text-to-motion models excel at efficient human motion generation, but existing approaches lack fine-grained controllability over the generation process. Consequently, modifying subtle postures within a motion or inserting new actions at specific moments remains a challenge, limiting the applicability of these methods in diverse scenarios. In light of these challenges, we introduce CoMo, a Controllable Motion generation model, adept at accurately generating and editing motions by leveraging the knowledge priors of large language models (LLMs). Specifically, CoMo decomposes motions into discrete and semantically meaningful pose codes, with each code encapsulating the semantics of a body part, representing elementary information such as "left knee slightly bent". Given textual inputs, CoMo autoregressively generates sequences of pose codes, which are then decoded into 3D motions. Leveraging pose codes as interpretable representations, an LLM can directly intervene in motion editing by adjusting the pose codes according to editing instructions. Experiments demonstrate that CoMo achieves competitive performance in motion generation compared to state-of-the-art models while, in human studies, CoMo substantially surpasses previous work in motion editing abilities.

Swetha Sirnam, Jinyu Yang, Tal Neiman, Mamshad Nayeem Rizve, Son Tran, Benjamin Yao, Trishul A Chilimbi, Shah Mubarak

Recent advancements in Multimodal Large Language Models (MLLMs) have revolutionized the field of vision-language understanding by incorporating visual perceptioning capabilities into Large Language Models (LLMs). The prevailing trend in this field involves the utilization of a vision encoder derived from vision-language contrastive learning (CL), showing expertise in capturing overall representations while facing difficulties in capturing detailed local patterns. In this work, we focus on enhancing the visual representations for MLLMs by combining high-frequency and fine-grained representations, obtained through masked image modeling (MIM), with semantically-enriched low-frequency representations captured by CL. To achieve this goal, we introduce X-Former which is a lightweight transformer module designed to exploit the complementary strengths of CL and MIM through an innovative interaction mechanism. Specifically, X-Former first bootstraps vision-language representation learning and multimodal-to-multimodal generative learning from two frozen vision encoders, i.e., CLIP-ViT (CL-based) \cite{radford2021clip} and MAE-ViT (MIM-based) \cite{he2022mae}. It further bootstraps vision-to-language generative learning from a frozen LLM to ensure visual features from X-Former can be interpreted by the LLM. To demonstrate the effectiveness of our approach, we assess its performance on tasks demanding fine-grained visual understanding. Our extensive empirical evaluations indicate that X-Former excels in visual reasoning tasks encompassing both structural and semantic categories within the GQA dataset. Assessment on a fine-grained visual perception benchmark further confirms its superior capabilities in visual understanding.

Pengxiang Ding, Han Zhao, Wenjie Zhang, Wenxuan Song, Min Zhang, Siteng Huang, Ningxi Yang, Donglin Wang

The important manifestation of robot intelligence is the ability to naturally interact and autonomously make decisions. Traditional approaches to robot control often compartmentalize perception, planning, and decision-making, simplifying system design but limiting the synergy between different information streams. This compartmentalization poses challenges in achieving seamless autonomous reasoning, decision-making, and action execution. To address these limitations, a novel paradigm, named Vision-Language-Action tasks for QUAdruped Robots (QUAR-VLA), has been introduced in this paper. This approach tightly integrates visual information and instructions to generate executable actions, effectively merging perception, planning, and decision-making. The central idea is to elevate the overall intelligence of the robot. Within this framework, a notable challenge lies in aligning fine-grained instructions with visual perception information. This emphasizes the complexity involved in ensuring that the robot accurately interprets and acts upon detailed instructions in harmony with its visual observations. Consequently, we propose QUAdruped Robotic Transformer (QUART), a VLA model to integrate visual information and instructions from diverse modalities as input and generates executable actions for real-world robots and present QUAdruped Robot Dataset (QUARD), a large-scale multi-task dataset including navigation, complex terrain locomotion, and whole-body manipulation tasks for training QUART models. Our extensive evaluation shows that our approach leads to performant robotic policies and enables QUART to obtain a range of generalization capabilities.

Guansong Lu, Yuanfan Guo, Jianhua Han, Minzhe Niu, Yihan Zeng, Xu Songcen, Zeyi Huang, Zhao Zhong, Wei Zhang, Hang Xu

Current large-scale diffusion models represent a giant leap forward in conditional image synthesis, capable of interpreting diverse cues like text, human poses, and edges. However, their reliance on substantial computational resources and extensive data collection remains a bottleneck. On the other hand, the integration of existing diffusion models, each specialized for different controls and operating in unique latent spaces, poses a challenge due to incompatible image resolutions and latent space embedding structures, hindering their joint use. Addressing these constraints, we present PanGu-Draw, a novel latent diffusion model designed for resource-efficient text-to-image synthesis that adeptly accommodates multiple control signals. We first propose a resource-efficient Time-Decoupling Training Strategy, which splits the monolithic text-to-image model into structure and texture generators. Each generator is trained using a regimen that maximizes data utilization and computational efficiency, cutting data preparation by 48% and reducing training resources by 51%. Secondly, we introduce Coop-Diffusion, an algorithm that enables the cooperative use of various pre-trained diffusion models with different latent spaces and predefined resolutions within a unified denoising process. This allows for multi-control image synthesis at arbitrary resolutions without the necessity for additional data or retraining. Empirical validations of Pangu-Draw show its exceptional prowess in text-to-image and multi-control image generation, suggesting a promising direction for future model training efficiencies and generation versatility. The largest 5B T2I PanGu-Draw model will be released soon on the Ascend platform.

Oguzhan Fatih Kar, Alessio Tonioni, Petra Poklukar, Achin Kulshrestha, Amir Zamir, Federico Tombari

Vision-language models (VLMs) are typically composed of a vision encoder, e.g. CLIP, and a language model (LM) that interprets the encoded features to solve downstream tasks. Despite remarkable progress, VLMs are subject to several shortcomings due to the limited capabilities of vision encoders, e.g. ``blindness'' to certain image features, visual hallucination, etc. To address these issues, we study broadening of the visual encoding capabilities of VLMs. We first comprehensively benchmark several vision encoders with different inductive biases for solving VLM tasks. We observe that there is no single encoding configuration that consistently achieves top performance across different tasks, and encoders with different biases can perform surprisingly similarly. Motivated by this, we introduce a method, named BRAVE, that consolidates features from multiple frozen encoders into a more versatile representation that can be directly fed as the input to a frozen LM. BRAVE achieves state-of-the-art performance on a broad range of captioning and VQA benchmarks and significantly reduces the aforementioned issues of VLMs, while requiring a smaller number of trainable parameters than existing methods and having a more compressed representation. Our results highlight the potential of incorporating different visual biases for a more broad and contextualized visual understanding of VLMs.

Ansel Blume, Khanh Duy Nguyen, Zhenhailong Wang, Yangyi Chen, Michal Shlapentokh-Rothman, Xiaomeng Jin, Jeonghwan Kim, Zhen Zhu 0006, Jiateng Liu, Kuan-Hao Huang 等

We present MIRACLE, a system for online, interpretable visual concept and video action recognition. Through a chat interface, users query the recognition system with an uploaded image or video. For images, MIRACLE returns concept predictions from its structured knowledge base, justifying its predictions with heatmaps and natural language-based attribute detections. For videos, MIRACLE predicts an action and justifies its prediction with time varying entity-entity relations. With its ability to learn new concepts in an online, few-shot manner and its support of dynamic changes to its knowledge base, MIRACLE represents a step forward in interpretable multimodal learning systems.

Dan Wang 0011, Xinrui Cui

We propose Interpretable Neural Radiance Fields (InNeRF) for generalizable 3D scene representation and rendering. In contrast to previous image-based rendering, which used two independent working processes of pooling-based fusion and MLP-based rendering, our framework unifies source-view fusion and target-view rendering processes via an end-to-end interpretable Transformer-based network. InNeRF enables the investigation of deep relationships between the target-rendering view and source views that were previously neglected by pooling-based fusion and fragmented rendering procedures. As a result, InNeRF improves model interpretability by enhancing the shape and appearance consistency of a 3D scene in both the surrounding view space and the ray-cast space. For a query rendering 3D point, InNeRF integrates both its projected 2D pixels from the surrounding source views and its adjacent 3D points along the query ray and simultaneously decodes this information into the query 3D point representation. Experiments show that InNeRF outperforms state-of-the-art image-based neural rendering methods in both scene-agnostic and per-scene finetuning scenarios, especially when there is a considerable disparity between source views and rendering views. The interpretation experiment shows that InNeRF can explain a query rendering process.

Shouyu Chen, Liang Hu 0004, Tangwei Ye, Zhongyuan Lai, Qi Zhang 0020, Ke Liu, Usman Naseem, Ke Sun, Nengjun Zhu

Interpretable and robust medical diagnoses are essential traits for practicing clinicians. Most computer-augmented diagnostic systems suffer from three major problems: non-interpretability, limited modality analysis, and narrow focus. Existing frameworks can either deal with multimodality to some extent but suffer from non-interpretability or partially interpretable but provide a limited modality and multifaceted capabilities. Our work aims to integrate all these aspects in one complete framework to fully utilize the full spectrum of information offered by multiple modalities and facets. We propose our solution via our novel architecture VR-DiagNet, consisting of a planner and a classifier, optimized iteratively and cohesively. VR-DiagNet simulates the perceptual process of clinicians via the use of volumetric imaging information integrated with radiomic features modality; at the same time, it recreates human thought processes via a customized Monte Carlo Tree Search (MCTS) which constructs a volume-tailored experience tree to identify slices of interest (SoIs) in our multi-slice perception space. We conducted extensive experiments across two diagnostic tasks comprising six public medical volumetric benchmark datasets. Our findings showcase superior performance, as evidenced by heightened accuracy and area under the curve (AUC) metrics, reduced computational overhead, and expedited convergence while conclusively illustrating the immense value of integrating volumetric and radiomic modalities for our current problem setup.

Yanghao Su, Jie Zhang 0073, Ting Xu 0004, Tianwei Zhang 0004, Weiming Zhang 0001, Nenghai Yu

Backdoor attacks pose a significant security vulnerability for deep neural networks (DNNs), enabling them to operate normally on clean inputs but manipulate predictions when specific trigger patterns occur. In this paper, we consider a practical post-training scenario backdoor defense, where the defender aims to evaluate whether a trained model has been compromised by backdoor attacks. Currently, post-training backdoor detection approaches often operate under the assumption that the defender has knowledge of the attack information, logit output from the model, and knowledge of the model parameters, limiting their implementation in practical scenarios. In contrast, our approach functions as a lightweight diagnostic scanning tool offering interpretability and visualization. By accessing the model to obtain hard labels, we construct decision boundaries within the convex combination of three samples. We present an intriguing observation of two phenomena in backdoored models: a noticeable shrinking of areas dominated by clean samples and a significant increase in the surrounding areas dominated by target labels. Leveraging this observation, we propose Model X-ray, a novel backdoor detection approach based on the analysis of illustrated two-dimensional (2D) decision boundaries. Our approach includes two strategies focused on the decision areas dominated by clean samples and the concentration of label distribution, and it can not only identify whether the target model is infected but also determine the target attacked label under the all-to-one attack strategy. Importantly, it accomplishes this solely by the predicted hard labels of clean inputs, regardless of any assumptions about attacks and prior knowledge of the training details of the model. Extensive experiments demonstrated that Model X-ray has outstanding effectiveness and efficiency across diverse backdoor attacks, datasets, and architectures. Besides, ablation studies on hyperparameters and more attack strategies and discussions are also provided.

Yalan Qin, Li Qian 0001

Multi-view clustering methods have been extensively explored in the last decades. This kind of methods is built on the assumption that the data are sampled from multiple subspaces with low dimension and each group fits into one of these subspaces. The quadratic or cubic computation complexity produced by these methods is inevitable, resulting in the difficulty for clustering multi-view datasets with large scales. Some efforts have been presented to select key anchors beforehand to capture the data distributions in different views. Despite significant progress, these methods pay few attentions to deriving provably scalable and correct method for finding the optimal shared anchor graph from the geometric interpretation perspective. They also ignore to give a well balance between the connectedness and subspace preserving properties of the shared anchor graph. In this paper, we propose the Fast Elastic- Net Multi-view Clustering (FENMC) from a geometric interpretation perspective. We provide the geometric analysis in determining the optimal shared anchor graph based on the introduced elastic-net regularizer for fast multi-view clustering, where the elastic-net regularizer is built on the mixture of L_2 and L_1 norms. We also give a theoretical justification for the balance between the connectedness and subspace preserving properties of the shared anchor graph for multi-view clustering. Our experiments on different datasets show that the proposed method not only obtains the satisfied clustering performance, but also deals with large-scale datasets with high efficiency.