Vision Transformers (ViTs) have become ubiquitous in computer vision. Despite their success, ViTs lack inductive biases, which can make it difficult to train them with limited data. To address this challenge, prior studies suggest training ViTs with self-supervised learning (SSL) and fine-tuning sequentially. However, we observe that jointly optimizing ViTs for the primary task and a Self-Supervised Auxiliary Task (SSAT) is surprisingly beneficial when the amount of training data is limited. We explore the appropriate SSL tasks that can be optimized alongside the primary task, the training schemes for these tasks, and the data scale at which they can be most effective. Our findings reveal that SSAT is a powerful technique that enables ViTs to leverage the unique characteristics of both the self-supervised and primary tasks, achieving better performance than typical ViTs pre-training with SSL and fine-tuning sequentially. Our experiments, conducted on 10 datasets, demonstrate that SSAT significantly improves ViT performance while reducing carbon footprint. We also confirm the effectiveness of SSAT in the video domain for deepfake detection, showcasing its generalizability. Our code is available at https://github.com/dominickrei/Limited-data-vits.
In this paper, we critically evaluate the capabilities of the state-of-the-art multimodal large language model, i.e., GPT-4 with Vision (GPT-4V), on Visual Question Answering (VQA) task. Our experiments thoroughly assess GPT-4V's proficiency in answering questions paired with images using both pathology and radiology datasets from 11 modalities (e.g. Microscopy, Dermoscopy, X-ray, CT, etc.) and fifteen objects of interests (brain, liver, lung, etc.). Our datasets encompass a comprehensive range of medical inquiries, including sixteen distinct question types. Throughout our evaluations, we devised textual prompts for GPT-4V, directing it to synergize visual and textual information. The experiments with accuracy score conclude that the current version of GPT-4V is not recommended for real-world diagnostics due to its unreliable and suboptimal accuracy in responding to diagnostic medical questions. In addition, we delineate seven unique facets of GPT-4V's behavior in medical VQA, highlighting its constraints within this complex arena. The complete details of our evaluation cases are accessible at https://github.com/ZhilingYan/GPT4V-Medical-Report.
In this work, we present 3DCoMPaT$^{++}$, a multimodal 2D/3D dataset with 160 million rendered views of more than 10 million stylized 3D shapes carefully annotated at the part-instance level, alongside matching RGB point clouds, 3D textured meshes, depth maps, and segmentation masks. 3DCoMPaT$^{++}$ covers 41 shape categories, 275 fine-grained part categories, and 293 fine-grained material classes that can be compositionally applied to parts of 3D objects. We render a subset of one million stylized shapes from four equally spaced views as well as four randomized views, leading to a total of 160 million renderings. Parts are segmented at the instance level, with coarse-grained and fine-grained semantic levels. We introduce a new task, called Grounded CoMPaT Recognition (GCR), to collectively recognize and ground compositions of materials on parts of 3D objects. Additionally, we report the outcomes of a data challenge organized at CVPR2023, showcasing the winning method's utilization of a modified PointNet$^{++}$ model trained on 6D inputs, and exploring alternative techniques for GCR enhancement. We hope our work will help ease future research on compositional 3D Vision.
Graph Neural Networks (GNNs) are powerful in learning semantics of graph data. Recently, a new paradigm "pre-train, prompt" has shown promising results in adapting GNNs to various tasks with less supervised data. The success of such paradigm can be attributed to the more consistent objectives of pre-training and task-oriented prompt tuning, where the pre-trained knowledge can be effectively transferred to downstream tasks. However, an overlooked issue of existing studies is that the structure information of graph is usually exploited during pre-training for learning node representations, while neglected in the prompt tuning stage for learning task-specific parameters. To bridge this gap, we propose a novel structure-based prompting method for GNNs, namely SAP, which consistently exploits structure information in both pre-training and prompt tuning stages. In particular, SAP 1) employs a dual-view contrastive learning to align the latent semantic spaces of node attributes and graph structure, and 2) incorporates structure information in prompted graph to elicit more pre-trained knowledge in prompt tuning. We conduct extensive experiments on node classification and graph classification tasks to show the effectiveness of SAP. Moreover, we show that SAP can lead to better performance in more challenging few-shot scenarios on both homophilous and heterophilous graphs.
Large language models have shown their remarkable capabilities as a general interface for various language-related applications. Motivated by this, we target to build a unified interface for completing many vision-language tasks including image description, visual question answering, and visual grounding, among others. The challenge is to use a single model for performing diverse vision-language tasks effectively with simple multi-modal instructions. Towards this objective, we introduce MiniGPT-v2, a model that can be treated as a unified interface for better handling various vision-language tasks. We propose using unique identifiers for different tasks when training the model. These identifiers enable our model to better distinguish each task instruction effortlessly and also improve the model learning efficiency for each task. After the three-stage training, the experimental results show that MiniGPT-v2 achieves strong performance on many visual question-answering and visual grounding benchmarks compared to other vision-language generalist models. Our model and codes are available at https://minigpt-v2.github.io/
Label noise is a common challenge in large datasets, as it can significantly degrade the generalization ability of deep neural networks. Most existing studies focus on noisy labels in computer vision; however, graph models encompass both node features and graph topology as input, and become more susceptible to label noise through message-passing mechanisms. Recently, only a few works have been proposed to tackle the label noise on graphs. One major limitation is that they assume the graph is homophilous and the labels are smoothly distributed. Nevertheless, real-world graphs may contain varying degrees of heterophily or even be heterophily-dominated, leading to the inadequacy of current methods. In this paper, we study graph label noise in the context of arbitrary heterophily, with the aim of rectifying noisy labels and assigning labels to previously unlabeled nodes. We begin by conducting two empirical analyses to explore the impact of graph homophily on graph label noise. Following observations, we propose a simple yet efficient algorithm, denoted as LP4GLN. Specifically, LP4GLN is an iterative algorithm with three steps: (1) reconstruct the graph to recover the homophily property, (2) utilize label propagation to rectify the noisy labels, (3) select high-confidence labels to retain for the next iteration. By iterating these steps, we obtain a set of correct labels, ultimately achieving high accuracy in the node classification task. The theoretical analysis is also provided to demonstrate its remarkable denoising "effect". Finally, we conduct experiments on 10 benchmark datasets under varying graph heterophily levels and noise types, comparing the performance of LP4GLN with 7 typical baselines. Our results illustrate the superior performance of the proposed LP4GLN.
Chain-of-Thought (CoT) prompting has proven to be effective in enhancing the reasoning capabilities of Large Language Models (LLMs) with at least 100 billion parameters. However, it is ineffective or even detrimental when applied to reasoning tasks in Smaller Language Models (SLMs) with less than 10 billion parameters. To address this limitation, we introduce Dialogue-guided Chain-of-Thought (DialCoT) which employs a dialogue format to generate intermediate reasoning steps, guiding the model toward the final answer. Additionally, we optimize the model's reasoning path selection using the Proximal Policy Optimization (PPO) algorithm, further enhancing its reasoning capabilities. Our method offers several advantages compared to previous approaches. Firstly, we transform the process of solving complex reasoning questions by breaking them down into a series of simpler sub-questions, significantly reducing the task difficulty and making it more suitable for SLMs. Secondly, we optimize the model's reasoning path selection through the PPO algorithm. We conduct comprehensive experiments on four arithmetic reasoning datasets, demonstrating that our method achieves significant performance improvements compared to state-of-the-art competitors.
Autonomous robotic ultrasound System (RUSS) has been extensively studied. However, fully automated ultrasound image acquisition is still challenging, partly due to the lack of study in combining two phases of path planning: guiding the ultrasound probe to the scan target and covering the scan surface or volume. This paper presents a system of Automated Path Planning for RUSS (APP-RUSS). Our focus is on the first phase of automation, which emphasizes directing the ultrasound probe's path toward the target over extended distances. Specifically, our APP-RUSS system consists of a RealSense D405 RGB-D camera that is employed for visual guidance of the UR5e robotic arm and a cubic Bezier curve path planning model that is customized for delivering the probe to the recognized target. APP-RUSS can contribute to understanding the integration of the two phases of path planning in robotic ultrasound imaging, paving the way for its clinical adoption.
We present VR PreM+, an innovative VR system designed to enhance web exploration beyond traditional computer screens. Unlike static 2D displays, VR PreM+ leverages 3D environments to create an immersive pre-learning experience. Using keyword-based information retrieval allows users to manage and connect various content sources in a dynamic 3D space, improving communication and data comparison. We conducted preliminary and user studies that demonstrated efficient information retrieval, increased user engagement, and a greater sense of presence. These findings yielded three design guidelines for future VR information systems: display, interaction, and user-centric design. VR PreM+ bridges the gap between traditional web browsing and immersive VR, offering an interactive and comprehensive approach to information acquisition. It holds promise for research, education, and beyond.
A fundamental limitation of object detectors is that they suffer from "spatial bias", and in particular perform less satisfactorily when detecting objects near image borders. For a long time, there has been a lack of effective ways to measure and identify spatial bias, and little is known about where it comes from and what degree it is. To this end, we present a new zone evaluation protocol, extending from the traditional evaluation to a more generalized one, which measures the detection performance over zones, yielding a series of Zone Precisions (ZPs). For the first time, we provide numerical results, showing that the object detectors perform quite unevenly across the zones. Surprisingly, the detector's performance in the 96\% border zone of the image does not reach the AP value (Average Precision, commonly regarded as the average detection performance in the entire image zone). To better understand spatial bias, a series of heuristic experiments are conducted. Our investigation excludes two intuitive conjectures about spatial bias that the object scale and the absolute positions of objects barely influence the spatial bias. We find that the key lies in the human-imperceptible divergence in data patterns between objects in different zones, thus eventually forming a visible performance gap between the zones. With these findings, we finally discuss a future direction for object detection, namely, spatial disequilibrium problem, aiming at pursuing a balanced detection ability over the entire image zone. By broadly evaluating 10 popular object detectors and 5 detection datasets, we shed light on the spatial bias of object detectors. We hope this work could raise a focus on detection robustness. The source codes, evaluation protocols, and tutorials are publicly available at \url{https://github.com/Zzh-tju/ZoneEval}.