Abstract:In this work, we introduce Libra, a prototype model with a decoupled vision system on a large language model (LLM). The decoupled vision system decouples inner-modal modeling and cross-modal interaction, yielding unique visual information modeling and effective cross-modal comprehension. Libra is trained through discrete auto-regressive modeling on both vision and language inputs. Specifically, we incorporate a routed visual expert with a cross-modal bridge module into a pretrained LLM to route the vision and language flows during attention computing to enable different attention patterns in inner-modal modeling and cross-modal interaction scenarios. Experimental results demonstrate that the dedicated design of Libra achieves a strong MLLM baseline that rivals existing works in the image-to-text scenario with merely 50 million training data, providing a new perspective for future multimodal foundation models. Code is available at https://github.com/YifanXu74/Libra.
Abstract:Visual grounding, which aims to ground a visual region via natural language, is a task that heavily relies on cross-modal alignment. Existing works utilized uni-modal pre-trained models to transfer visual/linguistic knowledge separately while ignoring the multimodal corresponding information. Motivated by recent advancements in contrastive language-image pre-training and low-rank adaptation (LoRA) methods, we aim to solve the grounding task based on multimodal pre-training. However, there exists significant task gaps between pre-training and grounding. Therefore, to address these gaps, we propose a concise and efficient hierarchical multimodal fine-grained modulation framework, namely HiVG. Specifically, HiVG consists of a multi-layer adaptive cross-modal bridge and a hierarchical multimodal low-rank adaptation (Hi LoRA) paradigm. The cross-modal bridge can address the inconsistency between visual features and those required for grounding, and establish a connection between multi-level visual and text features. Hi LoRA prevents the accumulation of perceptual errors by adapting the cross-modal features from shallow to deep layers in a hierarchical manner. Experimental results on five datasets demonstrate the effectiveness of our approach and showcase the significant grounding capabilities as well as promising energy efficiency advantages. The project page: https://github.com/linhuixiao/HiVG.
Abstract:In this paper, we for the first time explore helpful multi-modal contextual knowledge to understand novel categories for open-vocabulary object detection (OVD). The multi-modal contextual knowledge stands for the joint relationship across regions and words. However, it is challenging to incorporate such multi-modal contextual knowledge into OVD. The reason is that previous detection frameworks fail to jointly model multi-modal contextual knowledge, as object detectors only support vision inputs and no caption description is provided at test time. To this end, we propose a multi-modal contextual knowledge distillation framework, MMC-Det, to transfer the learned contextual knowledge from a teacher fusion transformer with diverse multi-modal masked language modeling (D-MLM) to a student detector. The diverse multi-modal masked language modeling is realized by an object divergence constraint upon traditional multi-modal masked language modeling (MLM), in order to extract fine-grained region-level visual contexts, which are vital to object detection. Extensive experiments performed upon various detection datasets show the effectiveness of our multi-modal context learning strategy, where our approach well outperforms the recent state-of-the-art methods.
Abstract:We introduce MQ-Det, an efficient architecture and pre-training strategy design to utilize both textual description with open-set generalization and visual exemplars with rich description granularity as category queries, namely, Multi-modal Queried object Detection, for real-world detection with both open-vocabulary categories and various granularity. MQ-Det incorporates vision queries into existing well-established language-queried-only detectors. A plug-and-play gated class-scalable perceiver module upon the frozen detector is proposed to augment category text with class-wise visual information. To address the learning inertia problem brought by the frozen detector, a vision conditioned masked language prediction strategy is proposed. MQ-Det's simple yet effective architecture and training strategy design is compatible with most language-queried object detectors, thus yielding versatile applications. Experimental results demonstrate that multi-modal queries largely boost open-world detection. For instance, MQ-Det significantly improves the state-of-the-art open-set detector GLIP by +7.8% zero-shot AP on the LVIS benchmark and averagely +6.3% AP on 13 few-shot downstream tasks, with merely 3% pre-training time required by GLIP. Code is available at https://github.com/YifanXu74/MQ-Det.
Abstract:Visual Grounding (VG) refers to locating a region described by expressions in a specific image, which is a critical topic in vision-language fields. To alleviate the dependence on labeled data, existing unsupervised methods try to locate regions using task-unrelated pseudo-labels. However, a large proportion of pseudo-labels are noisy and diversity scarcity in language taxonomy. Inspired by the advances in V-L pretraining, we consider utilizing the VLP models to realize unsupervised transfer learning in downstream grounding task. Thus, we propose CLIP-VG, a novel method that can conduct self-paced curriculum adapting of CLIP via exploiting pseudo-language labels to solve VG problem. By elaborating an efficient model structure, we first propose a single-source and multi-source curriculum adapting method for unsupervised VG to progressively sample more reliable cross-modal pseudo-labels to obtain the optimal model, thus achieving implicit knowledge exploiting and denoising. Our method outperforms the existing state-of-the-art unsupervised VG method Pseudo-Q in both single-source and multi-source scenarios with a large margin, i.e., 6.78%~10.67% and 11.39%~24.87% on RefCOCO/+/g datasets, even outperforms existing weakly supervised methods. The code and models will be released at \url{https://github.com/linhuixiao/CLIP-VG}.
Abstract:Although significant progress has been made in few-shot learning, most of existing few-shot learning methods require supervised pre-training on a large amount of samples of base classes, which limits their generalization ability in real world application. Recently, large-scale self-supervised vision-language models (e.g., CLIP) have provided a new paradigm for transferable visual representation learning. However, the pre-trained VLPs may neglect detailed visual information that is difficult to describe by language sentences, but important for learning an effective classifier in few-shot classification. To address the above problem, we propose a new framework, named Semantic-guided Visual Adapting (SgVA), which can effectively extend vision-language pre-trained models to produce discriminative task-specific visual features by comprehensively using a vision-specific contrastive loss, a cross-modal contrastive loss, and an implicit knowledge distillation. The implicit knowledge distillation is designed to transfer the fine-grained cross-modal knowledge to guide the updating of the vision adapter. State-of-the-art results on 13 datasets demonstrate that the adapted visual features can well complement the cross-modal features to improve few-shot image classification.
Abstract:Visual grounding focuses on establishing fine-grained alignment between vision and natural language, which has essential applications in multimodal reasoning systems. Existing methods use pre-trained query-agnostic visual backbones to extract visual feature maps independently without considering the query information. We argue that the visual features extracted from the visual backbones and the features really needed for multimodal reasoning are inconsistent. One reason is that there are differences between pre-training tasks and visual grounding. Moreover, since the backbones are query-agnostic, it is difficult to completely avoid the inconsistency issue by training the visual backbone end-to-end in the visual grounding framework. In this paper, we propose a Query-modulated Refinement Network (QRNet) to address the inconsistent issue by adjusting intermediate features in the visual backbone with a novel Query-aware Dynamic Attention (QD-ATT) mechanism and query-aware multiscale fusion. The QD-ATT can dynamically compute query-dependent visual attention at the spatial and channel levels of the feature maps produced by the visual backbone. We apply the QRNet to an end-to-end visual grounding framework. Extensive experiments show that the proposed method outperforms state-of-the-art methods on five widely used datasets.
Abstract:Graph convolutional networks (GCNs) based methods have achieved advanced performance on skeleton-based action recognition task. However, the skeleton graph cannot fully represent the motion information contained in skeleton data. In addition, the topology of the skeleton graph in the GCN-based methods is manually set according to natural connections, and it is fixed for all samples, which cannot well adapt to different situations. In this work, we propose a novel dynamic hypergraph convolutional networks (DHGCN) for skeleton-based action recognition. DHGCN uses hypergraph to represent the skeleton structure to effectively exploit the motion information contained in human joints. Each joint in the skeleton hypergraph is dynamically assigned the corresponding weight according to its moving, and the hypergraph topology in our model can be dynamically adjusted to different samples according to the relationship between the joints. Experimental results demonstrate that the performance of our model achieves competitive performance on three datasets: Kinetics-Skeleton 400, NTU RGB+D 60, and NTU RGB+D 120.
Abstract:Recently, the transductive graph-based methods have achieved great success in the few-shot classification task. However, most existing methods ignore exploring the class-level knowledge that can be easily learned by humans from just a handful of samples. In this paper, we propose an Explicit Class Knowledge Propagation Network (ECKPN), which is composed of the comparison, squeeze and calibration modules, to address this problem. Specifically, we first employ the comparison module to explore the pairwise sample relations to learn rich sample representations in the instance-level graph. Then, we squeeze the instance-level graph to generate the class-level graph, which can help obtain the class-level visual knowledge and facilitate modeling the relations of different classes. Next, the calibration module is adopted to characterize the relations of the classes explicitly to obtain the more discriminative class-level knowledge representations. Finally, we combine the class-level knowledge with the instance-level sample representations to guide the inference of the query samples. We conduct extensive experiments on four few-shot classification benchmarks, and the experimental results show that the proposed ECKPN significantly outperforms the state-of-the-art methods.
Abstract:Health management is getting increasing attention all over the world. However, existing health management mainly relies on hospital examination and treatment, which are complicated and untimely. The emerging of mobile devices provides the possibility to manage people's health status in a convenient and instant way. Estimation of health status can be achieved with various kinds of data streams continuously collected from wearable sensors. However, these data streams are multi-source and heterogeneous, containing complex temporal structures with local contextual and global temporal aspects, which makes the feature learning and data joint utilization challenging. We propose to model the behavior-related multi-source data streams with a local-global graph, which contains multiple local context sub-graphs to learn short term local context information with heterogeneous graph neural networks and a global temporal sub-graph to learn long term dependency with self-attention networks. Then health status is predicted based on the structure-aware representation learned from the local-global behavior graph. We take experiments on StudentLife dataset, and extensive results demonstrate the effectiveness of our proposed model.