Noisy annotations such as missing annotations and location shifts often exist in crowd counting datasets due to multi-scale head sizes, high occlusion, etc. These noisy annotations severely affect the model training, especially for density map-based methods. To alleviate the negative impact of noisy annotations, we propose a novel crowd counting model with one convolution head and one transformer head, in which these two heads can supervise each other in noisy areas, called Cross-Head Supervision. The resultant model, CHS-Net, can synergize different types of inductive biases for better counting. In addition, we develop a progressive cross-head supervision learning strategy to stabilize the training process and provide more reliable supervision. Extensive experimental results on ShanghaiTech and QNRF datasets demonstrate superior performance over state-of-the-art methods. Code is available at //github.com/RaccoonDML/CHSNet.
相關內容
To fully leverage the advantages of large-scale pre-trained language models (PLMs) on downstream tasks, it has become a ubiquitous adaptation paradigm to fine-tune the entire parameters of PLMs. However, this paradigm poses issues of inefficient updating and resource over-consuming for fine-tuning in data-scarce and resource-limited scenarios, because of the large scale of parameters in PLMs. To alleviate these concerns, in this paper, we propose a parameter-efficient fine-tuning method HiFi, that is, only the highly informative and strongly correlated attention heads for the specific task are fine-tuned. To search for those significant attention heads, we develop a novel framework to analyze the effectiveness of heads. Specifically, we first model the relationship between heads into a graph from two perspectives of information richness and correlation, and then apply PageRank algorithm to determine the relative importance of each head. Extensive experiments on the GLUE benchmark demonstrate the effectiveness of our method, and show that HiFi obtains state-of-the-art performance over the prior baselines.
Class-agnostic counting (CAC) aims to count objects of interest from a query image given few exemplars. This task is typically addressed by extracting the features of query image and exemplars respectively with (un)shared feature extractors and by matching their feature similarity, leading to an extract-\textit{then}-match paradigm. In this work, we show that CAC can be simplified in an extract-\textit{and}-match manner, particularly using a pretrained and plain vision transformer (ViT) where feature extraction and similarity matching are executed simultaneously within the self-attention. We reveal the rationale of such simplification from a decoupled view of the self-attention and point out that the simplification is only made possible if the query and exemplar tokens are concatenated as input. The resulting model, termed CACViT, simplifies the CAC pipeline and unifies the feature spaces between the query image and exemplars. In addition, we find CACViT naturally encodes background information within self-attention, which helps reduce background disturbance. Further, to compensate the loss of the scale and the order-of-magnitude information due to resizing and normalization in ViT, we present two effective strategies for scale and magnitude embedding. Extensive experiments on the FSC147 and the CARPK datasets show that CACViT significantly outperforms state-of-the-art CAC approaches in both effectiveness (23.60% error reduction) and generalization, which suggests CACViT provides a concise and strong baseline for CAC. Code will be available.
In-context learning is a new learning paradigm where a language model conditions on a few input-output pairs (demonstrations) and a test input, and directly outputs the prediction. It has been shown highly dependent on the provided demonstrations and thus promotes the research of demonstration retrieval: given a test input, relevant examples are retrieved from the training set to serve as informative demonstrations for in-context learning. While previous works focus on training task-specific retrievers for several tasks separately, these methods are often hard to transfer and scale on various tasks, and separately trained retrievers incur a lot of parameter storage and deployment cost. In this paper, we propose Unified Demonstration Retriever (\textbf{UDR}), a single model to retrieve demonstrations for a wide range of tasks. To train UDR, we cast various tasks' training signals into a unified list-wise ranking formulation by language model's feedback. Then we propose a multi-task list-wise ranking training framework, with an iterative mining strategy to find high-quality candidates, which can help UDR fully incorporate various tasks' signals. Experiments on 30+ tasks across 13 task families and multiple data domains show that UDR significantly outperforms baselines. Further analyses show the effectiveness of each proposed component and UDR's strong ability in various scenarios including different LMs (1.3B - 175B), unseen datasets, varying demonstration quantities, etc.
Robust watermarking tries to conceal information within a cover image/video imperceptibly that is resistant to various distortions. Recently, deep learning-based approaches for image watermarking have made significant advancements in robustness and invisibility. However, few studies focused on video watermarking using deep neural networks due to the high complexity and computational costs. Our paper aims to answer this research question: Can well-designed deep learning-based image watermarking be efficiently adapted to video watermarking? Our answer is positive. First, we revisit the workflow of deep learning-based watermarking methods that leads to a critical insight: temporal information in the video may be essential for general computer vision tasks but not for specific video watermarking. Inspired by this insight, we propose a method named ItoV for efficiently adapting deep learning-based Image watermarking to Video watermarking. Specifically, ItoV merges the temporal dimension of the video with the channel dimension to enable deep neural networks to treat videos as images. We further explore the effects of different convolutional blocks in video watermarking. We find that spatial convolution is the primary influential component in video watermarking and depthwise convolutions significantly reduce computational cost with negligible impact on performance. In addition, we propose a new frame loss to constrain that the watermark intensity in each video clip frame is consistent, significantly improving the invisibility. Extensive experiments show the superior performance of the adapted video watermarking method compared with the state-of-the-art methods on Kinetics-600 and Inter4K datasets, which demonstrate the efficacy of our method ItoV.
In this paper, we tackle two challenges in multimodal learning for visual recognition: 1) when missing-modality occurs either during training or testing in real-world situations; and 2) when the computation resources are not available to finetune on heavy transformer models. To this end, we propose to utilize prompt learning and mitigate the above two challenges together. Specifically, our modality-missing-aware prompts can be plugged into multimodal transformers to handle general missing-modality cases, while only requiring less than 1% learnable parameters compared to training the entire model. We further explore the effect of different prompt configurations and analyze the robustness to missing modality. Extensive experiments are conducted to show the effectiveness of our prompt learning framework that improves the performance under various missing-modality cases, while alleviating the requirement of heavy model re-training. Code is available.
Human-centric perception plays a vital role in vision and graphics. But their data annotations are prohibitively expensive. Therefore, it is desirable to have a versatile pre-train model that serves as a foundation for data-efficient downstream tasks transfer. To this end, we propose the Human-Centric Multi-Modal Contrastive Learning framework HCMoCo that leverages the multi-modal nature of human data (e.g. RGB, depth, 2D keypoints) for effective representation learning. The objective comes with two main challenges: dense pre-train for multi-modality data, efficient usage of sparse human priors. To tackle the challenges, we design the novel Dense Intra-sample Contrastive Learning and Sparse Structure-aware Contrastive Learning targets by hierarchically learning a modal-invariant latent space featured with continuous and ordinal feature distribution and structure-aware semantic consistency. HCMoCo provides pre-train for different modalities by combining heterogeneous datasets, which allows efficient usage of existing task-specific human data. Extensive experiments on four downstream tasks of different modalities demonstrate the effectiveness of HCMoCo, especially under data-efficient settings (7.16% and 12% improvement on DensePose Estimation and Human Parsing). Moreover, we demonstrate the versatility of HCMoCo by exploring cross-modality supervision and missing-modality inference, validating its strong ability in cross-modal association and reasoning.
Since the cyberspace consolidated as fifth warfare dimension, the different actors of the defense sector began an arms race toward achieving cyber superiority, on which research, academic and industrial stakeholders contribute from a dual vision, mostly linked to a large and heterogeneous heritage of developments and adoption of civilian cybersecurity capabilities. In this context, augmenting the conscious of the context and warfare environment, risks and impacts of cyber threats on kinetic actuations became a critical rule-changer that military decision-makers are considering. A major challenge on acquiring mission-centric Cyber Situational Awareness (CSA) is the dynamic inference and assessment of the vertical propagations from situations that occurred at the mission supportive Information and Communications Technologies (ICT), up to their relevance at military tactical, operational and strategical views. In order to contribute on acquiring CSA, this paper addresses a major gap in the cyber defence state-of-the-art: the dynamic identification of Key Cyber Terrains (KCT) on a mission-centric context. Accordingly, the proposed KCT identification approach explores the dependency degrees among tasks and assets defined by commanders as part of the assessment criteria. These are correlated with the discoveries on the operational network and the asset vulnerabilities identified thorough the supported mission development. The proposal is presented as a reference model that reveals key aspects for mission-centric KCT analysis and supports its enforcement and further enforcement by including an illustrative application case.
In many visual systems, visual tracking often bases on RGB image sequences, in which some targets are invalid in low-light conditions, and tracking performance is thus affected significantly. Introducing other modalities such as depth and infrared data is an effective way to handle imaging limitations of individual sources, but multi-modal imaging platforms usually require elaborate designs and cannot be applied in many real-world applications at present. Near-infrared (NIR) imaging becomes an essential part of many surveillance cameras, whose imaging is switchable between RGB and NIR based on the light intensity. These two modalities are heterogeneous with very different visual properties and thus bring big challenges for visual tracking. However, existing works have not studied this challenging problem. In this work, we address the cross-modal object tracking problem and contribute a new video dataset, including 654 cross-modal image sequences with over 481K frames in total, and the average video length is more than 735 frames. To promote the research and development of cross-modal object tracking, we propose a new algorithm, which learns the modality-aware target representation to mitigate the appearance gap between RGB and NIR modalities in the tracking process. It is plug-and-play and could thus be flexibly embedded into different tracking frameworks. Extensive experiments on the dataset are conducted, and we demonstrate the effectiveness of the proposed algorithm in two representative tracking frameworks against 17 state-of-the-art tracking methods. We will release the dataset for free academic usage, dataset download link and code will be released soon.
We study joint learning of Convolutional Neural Network (CNN) and Transformer for vision-language pre-training (VLPT) which aims to learn cross-modal alignments from millions of image-text pairs. State-of-the-art approaches extract salient image regions and align regions with words step-by-step. As region-based visual features usually represent parts of an image, it is challenging for existing vision-language models to fully understand the semantics from paired natural languages. In this paper, we propose SOHO to "See Out of tHe bOx" that takes a whole image as input, and learns vision-language representation in an end-to-end manner. SOHO does not require bounding box annotations which enables inference 10 times faster than region-based approaches. In particular, SOHO learns to extract comprehensive yet compact image features through a visual dictionary (VD) that facilitates cross-modal understanding. VD is designed to represent consistent visual abstractions of similar semantics. It is updated on-the-fly and utilized in our proposed pre-training task Masked Visual Modeling (MVM). We conduct experiments on four well-established vision-language tasks by following standard VLPT settings. In particular, SOHO achieves absolute gains of 2.0% R@1 score on MSCOCO text retrieval 5k test split, 1.5% accuracy on NLVR$^2$ test-P split, 6.7% accuracy on SNLI-VE test split, respectively.
Convolutional neural networks (CNNs) have shown dramatic improvements in single image super-resolution (SISR) by using large-scale external samples. Despite their remarkable performance based on the external dataset, they cannot exploit internal information within a specific image. Another problem is that they are applicable only to the specific condition of data that they are supervised. For instance, the low-resolution (LR) image should be a "bicubic" downsampled noise-free image from a high-resolution (HR) one. To address both issues, zero-shot super-resolution (ZSSR) has been proposed for flexible internal learning. However, they require thousands of gradient updates, i.e., long inference time. In this paper, we present Meta-Transfer Learning for Zero-Shot Super-Resolution (MZSR), which leverages ZSSR. Precisely, it is based on finding a generic initial parameter that is suitable for internal learning. Thus, we can exploit both external and internal information, where one single gradient update can yield quite considerable results. (See Figure 1). With our method, the network can quickly adapt to a given image condition. In this respect, our method can be applied to a large spectrum of image conditions within a fast adaptation process.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191