Pub Date : 2024-10-30DOI: 10.1007/s11263-024-02269-3
Emmanuel Hartman, Emery Pierson, Martin Bauer, Mohamed Daoudi, Nicolas Charon
This paper introduces a new framework for surface analysis derived from the general setting of elastic Riemannian metrics on shape spaces. Traditionally, those metrics are defined over the infinite dimensional manifold of immersed surfaces and satisfy specific invariance properties enabling the comparison of surfaces modulo shape preserving transformations such as reparametrizations. The specificity of our approach is to restrict the space of allowable transformations to predefined finite dimensional bases of deformation fields. These are estimated in a data-driven way so as to emulate specific types of surface transformations. This allows us to simplify the representation of the corresponding shape space to a finite dimensional latent space. However, in sharp contrast with methods involving e.g. mesh autoencoders, the latent space is equipped with a non-Euclidean Riemannian metric inherited from the family of elastic metrics. We demonstrate how this model can be effectively implemented to perform a variety of tasks on surface meshes which, importantly, does not assume these to be pre-registered or to even have a consistent mesh structure. We specifically validate our approach on human body shape and pose data as well as human face and hand scans for problems such as shape registration, interpolation, motion transfer or random pose generation.
{"title":"Basis Restricted Elastic Shape Analysis on the Space of Unregistered Surfaces","authors":"Emmanuel Hartman, Emery Pierson, Martin Bauer, Mohamed Daoudi, Nicolas Charon","doi":"10.1007/s11263-024-02269-3","DOIUrl":"https://doi.org/10.1007/s11263-024-02269-3","url":null,"abstract":"<p>This paper introduces a new framework for surface analysis derived from the general setting of elastic Riemannian metrics on shape spaces. Traditionally, those metrics are defined over the infinite dimensional manifold of immersed surfaces and satisfy specific invariance properties enabling the comparison of surfaces modulo shape preserving transformations such as reparametrizations. The specificity of our approach is to restrict the space of allowable transformations to predefined finite dimensional bases of deformation fields. These are estimated in a data-driven way so as to emulate specific types of surface transformations. This allows us to simplify the representation of the corresponding shape space to a finite dimensional latent space. However, in sharp contrast with methods involving e.g. mesh autoencoders, the latent space is equipped with a non-Euclidean Riemannian metric inherited from the family of elastic metrics. We demonstrate how this model can be effectively implemented to perform a variety of tasks on surface meshes which, importantly, does not assume these to be pre-registered or to even have a consistent mesh structure. We specifically validate our approach on human body shape and pose data as well as human face and hand scans for problems such as shape registration, interpolation, motion transfer or random pose generation.</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142556048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-28DOI: 10.1007/s11263-024-02279-1
Tianshan Liu, Kin-Man Lam, Bing-Kun Bao
As a crucial topic of high-level video understanding, weakly supervised online activity detection (WS-OAD) involves identifying the ongoing behaviors moment-to-moment in streaming videos, trained with solely cheap video-level annotations. It is essentially a challenging task, which requires addressing the entangled issues of the weakly supervised settings and online constraints. In this paper, we tackle the WS-OAD task from the knowledge-distillation (KD) perspective, which trains an online student detector to distill dual-level knowledge from a weakly supervised offline teacher model. To guarantee the completeness of knowledge transfer, we improve the vanilla KD framework from two aspects. First, we introduce an external memory bank to maintain the long-term activity prototypes, which serves as a bridge to align the activity semantics learned from the offline teacher and online student models. Second, to compensate the missing contexts of unseen near future, we leverage a curriculum learning paradigm to gradually train the online student detector to anticipate the future activity semantics. By dynamically scheduling the provided auxiliary future states, the online detector progressively distills contextual information from the offline model in an easy-to-hard course. Extensive experimental results on three public data sets demonstrate the superiority of our proposed method over the competing methods.
{"title":"A Memory-Assisted Knowledge Transferring Framework with Curriculum Anticipation for Weakly Supervised Online Activity Detection","authors":"Tianshan Liu, Kin-Man Lam, Bing-Kun Bao","doi":"10.1007/s11263-024-02279-1","DOIUrl":"https://doi.org/10.1007/s11263-024-02279-1","url":null,"abstract":"<p>As a crucial topic of high-level video understanding, weakly supervised online activity detection (WS-OAD) involves identifying the ongoing behaviors moment-to-moment in streaming videos, trained with solely cheap video-level annotations. It is essentially a challenging task, which requires addressing the entangled issues of the weakly supervised settings and online constraints. In this paper, we tackle the WS-OAD task from the knowledge-distillation (KD) perspective, which trains an online student detector to distill dual-level knowledge from a weakly supervised offline teacher model. To guarantee the completeness of knowledge transfer, we improve the vanilla KD framework from two aspects. First, we introduce an external memory bank to maintain the long-term activity prototypes, which serves as a bridge to align the activity semantics learned from the offline teacher and online student models. Second, to compensate the missing contexts of unseen near future, we leverage a curriculum learning paradigm to gradually train the online student detector to anticipate the future activity semantics. By dynamically scheduling the provided auxiliary future states, the online detector progressively distills contextual information from the offline model in an easy-to-hard course. Extensive experimental results on three public data sets demonstrate the superiority of our proposed method over the competing methods.</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142536604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-26DOI: 10.1007/s11263-024-02277-3
Guifang Zhang, Shijun Tan, Zhe Ji, Yuming Fang
Multimodal based person re-identification (ReID) has garnered increasing attention in recent years. However, the integration of visual and textual information encounters significant challenges. Biases in feature integration are frequently observed in existing methods, resulting in suboptimal performance and restricted generalization across a spectrum of ReID tasks. At the same time, since there is a domain gap between the datasets used by the pretraining model and the ReID datasets, it has a certain impact on the performance. In response to these challenges, we proposed a dynamic attention vision-language transformer network for the ReID task. In this network, a novel image-text dynamic attention module (ITDA) is designed to promote unbiased feature integration by dynamically assigning the importance of image and text representations. Additionally, an adapter module is adopted to address the domain gap between pretraining datasets and ReID datasets. Our network can capture complex connections between visual and textual information and achieve satisfactory performance. We conducted numerous experiments on ReID benchmarks to demonstrate the efficacy of our proposed method. The experimental results show that our method achieves state-of-the-art performance, surpassing existing integration strategies. These findings underscore the critical role of unbiased feature dynamic integration in enhancing the capabilities of multimodal based ReID models.
{"title":"Dynamic Attention Vision-Language Transformer Network for Person Re-identification","authors":"Guifang Zhang, Shijun Tan, Zhe Ji, Yuming Fang","doi":"10.1007/s11263-024-02277-3","DOIUrl":"https://doi.org/10.1007/s11263-024-02277-3","url":null,"abstract":"<p>Multimodal based person re-identification (ReID) has garnered increasing attention in recent years. However, the integration of visual and textual information encounters significant challenges. Biases in feature integration are frequently observed in existing methods, resulting in suboptimal performance and restricted generalization across a spectrum of ReID tasks. At the same time, since there is a domain gap between the datasets used by the pretraining model and the ReID datasets, it has a certain impact on the performance. In response to these challenges, we proposed a dynamic attention vision-language transformer network for the ReID task. In this network, a novel image-text dynamic attention module (ITDA) is designed to promote unbiased feature integration by dynamically assigning the importance of image and text representations. Additionally, an adapter module is adopted to address the domain gap between pretraining datasets and ReID datasets. Our network can capture complex connections between visual and textual information and achieve satisfactory performance. We conducted numerous experiments on ReID benchmarks to demonstrate the efficacy of our proposed method. The experimental results show that our method achieves state-of-the-art performance, surpassing existing integration strategies. These findings underscore the critical role of unbiased feature dynamic integration in enhancing the capabilities of multimodal based ReID models.</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142490658","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work focuses on generating high-quality images with specific style of reference images and content of provided textual descriptions. Current leading algorithms, i.e., DreamBooth and LoRA, require fine-tuning for each style, leading to time-consuming and computationally expensive processes. In this work, we propose StyleAdapter, a unified stylized image generation model capable of producing a variety of stylized images that match both the content of a given prompt and the style of reference images, without the need for per-style fine-tuning. It introduces a two-path cross-attention (TPCA) module to separately process style information and textual prompt, which cooperate with a semantic suppressing vision model (SSVM) to suppress the semantic content of style images. In this way, it can ensure that the prompt maintains control over the content of the generated images, while also mitigating the negative impact of semantic information in style references. This results in the content of the generated image adhering to the prompt, and its style aligning with the style references. Besides, our StyleAdapter can be integrated with existing controllable synthesis methods, such as T2I-adapter and ControlNet, to attain a more controllable and stable generation process. Extensive experiments demonstrate the superiority of our method over previous works.
{"title":"StyleAdapter: A Unified Stylized Image Generation Model","authors":"Zhouxia Wang, Xintao Wang, Liangbin Xie, Zhongang Qi, Ying Shan, Wenping Wang, Ping Luo","doi":"10.1007/s11263-024-02253-x","DOIUrl":"https://doi.org/10.1007/s11263-024-02253-x","url":null,"abstract":"<p>This work focuses on generating high-quality images with specific style of reference images and content of provided textual descriptions. Current leading algorithms, i.e., DreamBooth and LoRA, require fine-tuning for each style, leading to time-consuming and computationally expensive processes. In this work, we propose StyleAdapter, a unified stylized image generation model capable of producing a variety of stylized images that match both the content of a given prompt and the style of reference images, without the need for per-style fine-tuning. It introduces a two-path cross-attention (TPCA) module to separately process style information and textual prompt, which cooperate with a semantic suppressing vision model (SSVM) to suppress the semantic content of style images. In this way, it can ensure that the prompt maintains control over the content of the generated images, while also mitigating the negative impact of semantic information in style references. This results in the content of the generated image adhering to the prompt, and its style aligning with the style references. Besides, our StyleAdapter can be integrated with existing controllable synthesis methods, such as T2I-adapter and ControlNet, to attain a more controllable and stable generation process. Extensive experiments demonstrate the superiority of our method over previous works.</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142489489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-25DOI: 10.1007/s11263-024-02254-w
Jiyang Guan, Jian Liang, Yanbo Wang, Ran He
Face recognition has witnessed remarkable advancements in recent years, thanks to the development of deep learning techniques. However, an off-the-shelf face recognition model as a commercial service could be stolen by model stealing attacks, posing great threats to the rights of the model owner. Model fingerprinting, as a model stealing detection method, aims to verify whether a suspect model is stolen from the victim model, gaining more and more attention nowadays. Previous methods always utilize transferable adversarial examples as the model fingerprint, but this method is known to be sensitive to adversarial defense and transfer learning techniques. To address this issue, we consider the pairwise relationship between samples instead and propose a novel yet simple model stealing detection method based on SAmple Correlation (SAC). Specifically, we present SAC-JC that selects JPEG compressed samples as model inputs and calculates the correlation matrix among their model outputs. Extensive results validate that SAC successfully defends against various model stealing attacks in deep face recognition, encompassing face verification and face emotion recognition, exhibiting the highest performance in terms of AUC, p-value and F1 score. Furthermore, we extend our evaluation of SAC-JC to object recognition datasets including Tiny-ImageNet and CIFAR10, which also demonstrates the superior performance of SAC-JC to previous methods. The code will be available at https://github.com/guanjiyang/SAC_JC.
{"title":"Sample Correlation for Fingerprinting Deep Face Recognition","authors":"Jiyang Guan, Jian Liang, Yanbo Wang, Ran He","doi":"10.1007/s11263-024-02254-w","DOIUrl":"https://doi.org/10.1007/s11263-024-02254-w","url":null,"abstract":"<p>Face recognition has witnessed remarkable advancements in recent years, thanks to the development of deep learning techniques. However, an off-the-shelf face recognition model as a commercial service could be stolen by model stealing attacks, posing great threats to the rights of the model owner. Model fingerprinting, as a model stealing detection method, aims to verify whether a suspect model is stolen from the victim model, gaining more and more attention nowadays. Previous methods always utilize transferable adversarial examples as the model fingerprint, but this method is known to be sensitive to adversarial defense and transfer learning techniques. To address this issue, we consider the pairwise relationship between samples instead and propose a novel yet simple model stealing detection method based on SAmple Correlation (SAC). Specifically, we present SAC-JC that selects JPEG compressed samples as model inputs and calculates the correlation matrix among their model outputs. Extensive results validate that SAC successfully defends against various model stealing attacks in deep face recognition, encompassing face verification and face emotion recognition, exhibiting the highest performance in terms of AUC, <i>p</i>-value and F1 score. Furthermore, we extend our evaluation of SAC-JC to object recognition datasets including Tiny-ImageNet and CIFAR10, which also demonstrates the superior performance of SAC-JC to previous methods. The code will be available at https://github.com/guanjiyang/SAC_JC.</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142490657","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-24DOI: 10.1007/s11263-024-02271-9
David Junhao Zhang, Jay Zhangjie Wu, Jia-Wei Liu, Rui Zhao, Lingmin Ran, Yuchao Gu, Difei Gao, Mike Zheng Shou
Significant advancements have been achieved in the realm of large-scale pre-trained text-to-video Diffusion Models (VDMs). However, previous methods either rely solely on pixel-based VDMs, which come with high computational costs, or on latent-based VDMs, which often struggle with precise text-video alignment. In this paper, we are the first to propose a hybrid model, dubbed as Show-1, which marries pixel-based and latent-based VDMs for text-to-video generation. Our model first uses pixel-based VDMs to produce a low-resolution video of strong text-video correlation. After that, we propose a novel expert translation method that employs the latent-based VDMs to further upsample the low-resolution video to high resolution, which can also remove potential artifacts and corruptions from low-resolution videos. Compared to latent VDMs, Show-1 can produce high-quality videos of precise text-video alignment; Compared to pixel VDMs, Show-1 is much more efficient (GPU memory usage during inference is 15 G vs. 72 G). Furthermore, our Show-1 model can be readily adapted for motion customization and video stylization applications through simple temporal attention layer finetuning. Our model achieves state-of-the-art performance on standard video generation benchmarks. Code of Show-1 is publicly available and more videos can be found here.
{"title":"Show-1: Marrying Pixel and Latent Diffusion Models for Text-to-Video Generation","authors":"David Junhao Zhang, Jay Zhangjie Wu, Jia-Wei Liu, Rui Zhao, Lingmin Ran, Yuchao Gu, Difei Gao, Mike Zheng Shou","doi":"10.1007/s11263-024-02271-9","DOIUrl":"https://doi.org/10.1007/s11263-024-02271-9","url":null,"abstract":"<p>Significant advancements have been achieved in the realm of large-scale pre-trained text-to-video Diffusion Models (VDMs). However, previous methods either rely solely on pixel-based VDMs, which come with high computational costs, or on latent-based VDMs, which often struggle with precise text-video alignment. In this paper, we are the first to propose a hybrid model, dubbed as Show-1, which marries pixel-based and latent-based VDMs for text-to-video generation. Our model first uses pixel-based VDMs to produce a low-resolution video of strong text-video correlation. After that, we propose a novel expert translation method that employs the latent-based VDMs to further upsample the low-resolution video to high resolution, which can also remove potential artifacts and corruptions from low-resolution videos. Compared to latent VDMs, Show-1 can produce high-quality videos of precise text-video alignment; Compared to pixel VDMs, Show-1 is much more efficient (GPU memory usage during inference is 15 G vs. 72 G). Furthermore, our Show-1 model can be readily adapted for motion customization and video stylization applications through simple temporal attention layer finetuning. Our model achieves state-of-the-art performance on standard video generation benchmarks. Code of Show-1 is publicly available and more videos can be found here.</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142489488","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-22DOI: 10.1007/s11263-024-02251-z
Edoardo Mello Rella, Ajad Chhatkuli, Ender Konukoglu, Luc Van Gool
Neural implicit fields have recently shown increasing success in representing, learning and analysis of 3D shapes. Signed distance fields and occupancy fields are still the preferred choice of implicit representations with well-studied properties, despite their restriction to closed surfaces. With neural networks, unsigned distance fields as well as several other variations and training principles have been proposed with the goal to represent all classes of shapes. In this paper, we develop a novel and yet a fundamental representation considering unit vectors in 3D space and call it Vector Field (VF). At each point in (mathbb {R}^3), VF is directed to the closest point on the surface. We theoretically demonstrate that VF can be easily transformed to surface density by computing the flux density. Unlike other standard representations, VF directly encodes an important physical property of the surface, its normal. We further show the advantages of VF representation, in learning open, closed, or multi-layered surfaces. We show that, thanks to the continuity property of the neural optimization with VF, a separate distance field becomes unnecessary for extracting surfaces from the implicit field via Marching Cubes. We compare our method on several datasets including ShapeNet where the proposed new neural implicit field shows superior accuracy in representing any type of shape, outperforming other standard methods. Codes are available at https://github.com/edomel/ImplicitVF.
{"title":"Neural Vector Fields for Implicit Surface Representation and Inference","authors":"Edoardo Mello Rella, Ajad Chhatkuli, Ender Konukoglu, Luc Van Gool","doi":"10.1007/s11263-024-02251-z","DOIUrl":"https://doi.org/10.1007/s11263-024-02251-z","url":null,"abstract":"<p>Neural implicit fields have recently shown increasing success in representing, learning and analysis of 3D shapes. Signed distance fields and occupancy fields are still the preferred choice of implicit representations with well-studied properties, despite their restriction to closed surfaces. With neural networks, unsigned distance fields as well as several other variations and training principles have been proposed with the goal to represent all classes of shapes. In this paper, we develop a novel and yet a fundamental representation considering unit vectors in 3D space and call it Vector Field (VF). At each point in <span>(mathbb {R}^3)</span>, VF is directed to the closest point on the surface. We theoretically demonstrate that VF can be easily transformed to surface density by computing the flux density. Unlike other standard representations, VF directly encodes an important physical property of the surface, its normal. We further show the advantages of VF representation, in learning open, closed, or multi-layered surfaces. We show that, thanks to the continuity property of the neural optimization with VF, a separate distance field becomes unnecessary for extracting surfaces from the implicit field via Marching Cubes. We compare our method on several datasets including ShapeNet where the proposed new neural implicit field shows superior accuracy in representing any type of shape, outperforming other standard methods. Codes are available at https://github.com/edomel/ImplicitVF.\u0000</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142487455","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-22DOI: 10.1007/s11263-024-02202-8
Lucas Ventura, Cordelia Schmid, Gül Varol
We describe a protocol to study text-to-video retrieval training with unlabeled videos, where we assume (i) no access to labels for any videos, i.e., no access to the set of ground-truth captions, but (ii) access to labeled images in the form of text. Using image expert models is a realistic scenario given that annotating images is cheaper therefore scalable, in contrast to expensive video labeling schemes. Recently, zero-shot image experts such as CLIP have established a new strong baseline for video understanding tasks. In this paper, we make use of this progress and instantiate the image experts from two types of models: a text-to-image retrieval model to provide an initial backbone, and image captioning models to provide supervision signal into unlabeled videos. We show that automatically labeling video frames with image captioning allows text-to-video retrieval training. This process adapts the features to the target domain at no manual annotation cost, consequently outperforming the strong zero-shot CLIP baseline. During training, we sample captions from multiple video frames that best match the visual content, and perform a temporal pooling over frame representations by scoring frames according to their relevance to each caption. We conduct extensive ablations to provide insights and demonstrate the effectiveness of this simple framework by outperforming the CLIP zero-shot baselines on text-to-video retrieval on three standard datasets, namely ActivityNet, MSR-VTT, and MSVD. Code and models will be made publicly available.
{"title":"Learning Text-to-Video Retrieval from Image Captioning","authors":"Lucas Ventura, Cordelia Schmid, Gül Varol","doi":"10.1007/s11263-024-02202-8","DOIUrl":"https://doi.org/10.1007/s11263-024-02202-8","url":null,"abstract":"<p>We describe a protocol to study text-to-video retrieval training with unlabeled videos, where we assume (i) no access to labels for any videos, i.e., no access to the set of ground-truth captions, but (ii) access to labeled images in the form of text. Using image expert models is a realistic scenario given that annotating images is cheaper therefore scalable, in contrast to expensive video labeling schemes. Recently, zero-shot image experts such as CLIP have established a new strong baseline for video understanding tasks. In this paper, we make use of this progress and instantiate the image experts from two types of models: a text-to-image retrieval model to provide an initial backbone, and image captioning models to provide supervision signal into unlabeled videos. We show that automatically labeling video frames with image captioning allows text-to-video retrieval training. This process adapts the features to the target domain at no manual annotation cost, consequently outperforming the strong zero-shot CLIP baseline. During training, we sample captions from multiple video frames that best match the visual content, and perform a temporal pooling over frame representations by scoring frames according to their relevance to each caption. We conduct extensive ablations to provide insights and demonstrate the effectiveness of this simple framework by outperforming the CLIP zero-shot baselines on text-to-video retrieval on three standard datasets, namely ActivityNet, MSR-VTT, and MSVD. Code and models will be made publicly available.\u0000</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142487598","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-21DOI: 10.1007/s11263-024-02267-5
Zhongyang Zhu, Jie Tang
The state-of-the-art methods for story visualization demonstrate a significant demand for training data and storage, as well as limited flexibility in story presentation, thereby rendering them impractical for real-world applications. We introduce CogCartoon, a practical story visualization method based on pre-trained diffusion models. To alleviate dependence on data and storage, we propose an innovative strategy of character-plugin generation that can represent a specific character as a compact 316 KB plugin by using a few training samples. To facilitate enhanced flexibility, we employ a strategy of plugin-guided and layout-guided inference, enabling users to seamlessly incorporate new characters and custom layouts into the generated image results at their convenience. We have conducted comprehensive qualitative and quantitative studies, providing compelling evidence for the superiority of CogCartoon over existing methodologies. Moreover, CogCartoon demonstrates its power in tackling challenging tasks, including long story visualization and realistic style story visualization.
{"title":"CogCartoon: Towards Practical Story Visualization","authors":"Zhongyang Zhu, Jie Tang","doi":"10.1007/s11263-024-02267-5","DOIUrl":"https://doi.org/10.1007/s11263-024-02267-5","url":null,"abstract":"<p>The state-of-the-art methods for story visualization demonstrate a significant demand for training data and storage, as well as limited flexibility in story presentation, thereby rendering them impractical for real-world applications. We introduce CogCartoon, a practical story visualization method based on pre-trained diffusion models. To alleviate dependence on data and storage, we propose an innovative strategy of character-plugin generation that can represent a specific character as a compact 316 KB plugin by using a few training samples. To facilitate enhanced flexibility, we employ a strategy of plugin-guided and layout-guided inference, enabling users to seamlessly incorporate new characters and custom layouts into the generated image results at their convenience. We have conducted comprehensive qualitative and quantitative studies, providing compelling evidence for the superiority of CogCartoon over existing methodologies. Moreover, CogCartoon demonstrates its power in tackling challenging tasks, including long story visualization and realistic style story visualization.</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142451997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Few-shot Segmentation aims to segment the interested objects in the query image with just a handful of labeled samples (i.e., support images). Previous schemes would leverage the similarity between support-query pixel pairs to construct the pixel-level semantic correlation. However, in remote sensing scenarios with extreme intra-class variations and cluttered backgrounds, such pixel-level correlations may produce tremendous mismatches, resulting in semantic ambiguity between the query foreground (FG) and background (BG) pixels. To tackle this problem, we propose a novel Agent Mining Transformer, which adaptively mines a set of local-aware agents to construct agent-level semantic correlation. Compared with pixel-level semantics, the given agents are equipped with local-contextual information and possess a broader receptive field. At this point, different query pixels can selectively aggregate the fine-grained local semantics of different agents, thereby enhancing the semantic clarity between query FG and BG pixels. Concretely, the Agent Learning Encoder is first proposed to erect the optimal transport plan that arranges different agents to aggregate support semantics under different local regions. Then, for further optimizing the agents, the Agent Aggregation Decoder and the Semantic Alignment Decoder are constructed to break through the limited support set for mining valuable class-specific semantics from unlabeled data sources and the query image itself, respectively. Extensive experiments on the remote sensing benchmark iSAID indicate that the proposed method achieves state-of-the-art performance. Surprisingly, our method remains quite competitive when extended to more common natural scenarios, i.e., PASCAL-(5^i) and COCO-(20^{i}).
{"title":"AgMTR: Agent Mining Transformer for Few-Shot Segmentation in Remote Sensing","authors":"Hanbo Bi, Yingchao Feng, Yongqiang Mao, Jianning Pei, Wenhui Diao, Hongqi Wang, Xian Sun","doi":"10.1007/s11263-024-02252-y","DOIUrl":"https://doi.org/10.1007/s11263-024-02252-y","url":null,"abstract":"<p>Few-shot Segmentation aims to segment the interested objects in the query image with just a handful of labeled samples (i.e., support images). Previous schemes would leverage the similarity between support-query pixel pairs to construct the pixel-level semantic correlation. However, in remote sensing scenarios with extreme intra-class variations and cluttered backgrounds, such pixel-level correlations may produce tremendous mismatches, resulting in semantic ambiguity between the query foreground (FG) and background (BG) pixels. To tackle this problem, we propose a novel Agent Mining Transformer, which adaptively mines a set of local-aware agents to construct agent-level semantic correlation. Compared with pixel-level semantics, the given agents are equipped with local-contextual information and possess a broader receptive field. At this point, different query pixels can selectively aggregate the fine-grained local semantics of different agents, thereby enhancing the semantic clarity between query FG and BG pixels. Concretely, the Agent Learning Encoder is first proposed to erect the optimal transport plan that arranges different agents to aggregate support semantics under different local regions. Then, for further optimizing the agents, the Agent Aggregation Decoder and the Semantic Alignment Decoder are constructed to break through the limited support set for mining valuable class-specific semantics from unlabeled data sources and the query image itself, respectively. Extensive experiments on the remote sensing benchmark iSAID indicate that the proposed method achieves state-of-the-art performance. Surprisingly, our method remains quite competitive when extended to more common natural scenarios, i.e., PASCAL-<span>(5^i)</span> and COCO-<span>(20^{i})</span>.\u0000</p>","PeriodicalId":13752,"journal":{"name":"International Journal of Computer Vision","volume":null,"pages":null},"PeriodicalIF":19.5,"publicationDate":"2024-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142486936","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}