Publicacions CVC -- Query Results

	Publicacions CVC Home \| Show All \| Simple Search \| Advanced Search \| Add Record \| Import	Login Quick Search: Field: contains: ...
	2356–2370 of 3413 records found matching your query (RSS):

Search & Display Options

Select All Deselect All

[141–150] << 151 152 153 154 155 156 157 158 159 160 >> [161–170]

List View

Citations

Details

	Records
	Author	Dipam Goswami; J Schuster; Joost Van de Weijer; Didier Stricker
	Title	Attribution-aware Weight Transfer: A Warm-Start Initialization for Class-Incremental Semantic Segmentation			Type	Conference Article
	Year	2023	Publication	Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision	Abbreviated Journal
	Volume		Issue		Pages	3195-3204
	Keywords
	Abstract	Attribution-aware Weight Transfer: A Warm-Start Initialization for Class-Incremental Semantic Segmentation. D Goswami, R Schuster, J van de Weijer, D Stricker. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023, pp. 3195-3204
	Address	Waikoloa; Hawai; USA; January 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	WACV
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ GSW2023			Serial	3901
Permanent link to this record



	Author	Senmao Li; Joost Van de Weijer; Yaxing Wang; Fahad Shahbaz Khan; Meiqin Liu; Jian Yang
	Title	3D-Aware Multi-Class Image-to-Image Translation with NeRFs			Type	Conference Article
	Year	2023	Publication	36th IEEE Conference on Computer Vision and Pattern Recognition	Abbreviated Journal
	Volume		Issue		Pages	12652-12662
	Keywords
	Abstract	Recent advances in 3D-aware generative models (3D-aware GANs) combined with Neural Radiance Fields (NeRF) have achieved impressive results. However no prior works investigate 3D-aware GANs for 3D consistent multiclass image-to-image (3D-aware 121) translation. Naively using 2D-121 translation methods suffers from unrealistic shape/identity change. To perform 3D-aware multiclass 121 translation, we decouple this learning process into a multiclass 3D-aware GAN step and a 3D-aware 121 translation step. In the first step, we propose two novel techniques: a new conditional architecture and an effective training strategy. In the second step, based on the well-trained multiclass 3D-aware GAN architecture, that preserves view-consistency, we construct a 3D-aware 121 translation system. To further reduce the view-consistency problems, we propose several new techniques, including a U-net-like adaptor network design, a hierarchical representation constrain and a relative regularization loss. In exten-sive experiments on two datasets, quantitative and qualitative results demonstrate that we successfully perform 3D-aware 121 translation with multi-view consistency. Code is available in 3DI2I.
	Address	Vancouver; Canada; June 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	CVPR
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ LWW2023b			Serial	3920
Permanent link to this record



	Author	Albin Soutif; Antonio Carta; Joost Van de Weijer
	Title	Improving Online Continual Learning Performance and Stability with Temporal Ensembles			Type	Conference Article
	Year	2023	Publication	2nd Conference on Lifelong Learning Agents	Abbreviated Journal
	Volume		Issue		Pages
	Keywords
	Abstract	Neural networks are very effective when trained on large datasets for a large number of iterations. However, when they are trained on non-stationary streams of data and in an online fashion, their performance is reduced (1) by the online setup, which limits the availability of data, (2) due to catastrophic forgetting because of the non-stationary nature of the data. Furthermore, several recent works (Caccia et al., 2022; Lange et al., 2023) arXiv:2205.13452 showed that replay methods used in continual learning suffer from the stability gap, encountered when evaluating the model continually (rather than only on task boundaries). In this article, we study the effect of model ensembling as a way to improve performance and stability in online continual learning. We notice that naively ensembling models coming from a variety of training tasks increases the performance in online continual learning considerably. Starting from this observation, and drawing inspirations from semi-supervised learning ensembling methods, we use a lightweight temporal ensemble that computes the exponential moving average of the weights (EMA) at test time, and show that it can drastically increase the performance and stability when used in combination with several methods from the literature.
	Address	Montreal; Canada; August 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	COLLAS
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ SCW2023			Serial	3922
Permanent link to this record



	Author	Alejandro Ariza-Casabona; Bartlomiej Twardowski; Tri Kurniawan Wijaya
	Title	Exploiting Graph Structured Cross-Domain Representation for Multi-domain Recommendation			Type	Conference Article
	Year	2023	Publication	European Conference on Information Retrieval – ECIR 2023: Advances in Information Retrieval	Abbreviated Journal
	Volume	13980	Issue		Pages	49–65
	Keywords
	Abstract	Multi-domain recommender systems benefit from cross-domain representation learning and positive knowledge transfer. Both can be achieved by introducing a specific modeling of input data (i.e. disjoint history) or trying dedicated training regimes. At the same time, treating domains as separate input sources becomes a limitation as it does not capture the interplay that naturally exists between domains. In this work, we efficiently learn multi-domain representation of sequential users’ interactions using graph neural networks. We use temporal intra- and inter-domain interactions as contextual information for our method called MAGRec (short for Multi-dom Ain Graph-based Recommender). To better capture all relations in a multi-domain setting, we learn two graph-based sequential representations simultaneously: domain-guided for recent user interest, and general for long-term interest. This approach helps to mitigate the negative knowledge transfer problem from multiple domains and improve overall representation. We perform experiments on publicly available datasets in different scenarios where MAGRec consistently outperforms state-of-the-art methods. Furthermore, we provide an ablation study and discuss further extensions of our method.
	Address
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title	LNCS
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ECIR
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ ATK2023			Serial	3933
Permanent link to this record



	Author	Dipam Goswami; Yuyang Liu ; Bartlomiej Twardowski; Joost Van de Weijer
	Title	FeCAM: Exploiting the Heterogeneity of Class Distributions in Exemplar-Free Continual Learning			Type	Conference Article
	Year	2023	Publication	37th Annual Conference on Neural Information Processing Systems	Abbreviated Journal
	Volume		Issue		Pages
	Keywords
	Abstract	Poster
	Address	New Orleans; USA; December 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	NEURIPS
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ GLT2023			Serial	3934
Permanent link to this record



	Author	Kai Wang; Fei Yang; Shiqi Yang; Muhammad Atif Butt; Joost Van de Weijer
	Title	Dynamic Prompt Learning: Addressing Cross-Attention Leakage for Text-Based Image Editing			Type	Conference Article
	Year	2023	Publication	37th Annual Conference on Neural Information Processing Systems	Abbreviated Journal
	Volume		Issue		Pages
	Keywords
	Abstract	Poster
	Address	New Orleans; USA; December 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	NEURIPS
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ WYY2023			Serial	3935
Permanent link to this record



	Author	ChuanMing Fang; Kai Wang; Joost Van de Weijer
	Title	IterInv: Iterative Inversion for Pixel-Level T2I Models			Type	Conference Article
	Year	2023	Publication	37th Annual Conference on Neural Information Processing Systems	Abbreviated Journal
	Volume		Issue		Pages
	Keywords
	Abstract	Large-scale text-to-image diffusion models have been a ground-breaking development in generating convincing images following an input text prompt. The goal of image editing research is to give users control over the generated images by modifying the text prompt. Current image editing techniques are relying on DDIM inversion as a common practice based on the Latent Diffusion Models (LDM). However, the large pretrained T2I models working on the latent space as LDM suffer from losing details due to the first compression stage with an autoencoder mechanism. Instead, another mainstream T2I pipeline working on the pixel level, such as Imagen and DeepFloyd-IF, avoids this problem. They are commonly composed of several stages, normally with a text-to-image stage followed by several super-resolution stages. In this case, the DDIM inversion is unable to find the initial noise to generate the original image given that the super-resolution diffusion models are not compatible with the DDIM technique. According to our experimental findings, iteratively concatenating the noisy image as the condition is the root of this problem. Based on this observation, we develop an iterative inversion (IterInv) technique for this stream of T2I models and verify IterInv with the open-source DeepFloyd-IF model. By combining our method IterInv with a popular image editing method, we prove the application prospects of IterInv. The code will be released at \url{this https URL}.
	Address	New Orleans; USA; December 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	NEURIPS
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ FWW2023			Serial	3936
Permanent link to this record



	Author	Albin Soutif; Antonio Carta; Andrea Cossu; Julio Hurtado; Hamed Hemati; Vincenzo Lomonaco; Joost Van de Weijer
	Title	A Comprehensive Empirical Evaluation on Online Continual Learning			Type	Conference Article
	Year	2023	Publication	Visual Continual Learning (ICCV-W)	Abbreviated Journal
	Volume		Issue		Pages
	Keywords
	Abstract	Online continual learning aims to get closer to a live learning experience by learning directly on a stream of data with temporally shifting distribution and by storing a minimum amount of data from that stream. In this empirical evaluation, we evaluate various methods from the literature that tackle online continual learning. More specifically, we focus on the class-incremental setting in the context of image classification, where the learner must learn new classes incrementally from a stream of data. We compare these methods on the Split-CIFAR100 and Split-TinyImagenet benchmarks, and measure their average accuracy, forgetting, stability, and quality of the representations, to evaluate various aspects of the algorithm at the end but also during the whole training period. We find that most methods suffer from stability and underfitting issues. However, the learned representations are comparable to i.i.d. training under the same computational budget. No clear winner emerges from the results and basic experience replay, when properly tuned and implemented, is a very strong baseline. We release our modular and extensible codebase at this https URL based on the avalanche framework to reproduce our results and encourage future research.
	Address	Paris; France; October 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCVW
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ SCC2023			Serial	3938
Permanent link to this record



	Author	Matej Kristan; Jiri Matas; Martin Danelljan; Michael Felsberg; Hyung Jin Chang; Luka Cehovin Zajc; Alan Lukezic; Ondrej Drbohlav; Zhongqun Zhang; Khanh-Tung Tran; Xuan-Son Vu; Johanna Bjorklund; Christoph Mayer; Yushan Zhang; Lei Ke; Jie Zhao; Gustavo Fernandez; Noor Al-Shakarji; Dong An; Michael Arens; Stefan Becker; Goutam Bhat; Sebastian Bullinger; Antoni B. Chan; Shijie Chang; Hanyuan Chen; Xin Chen; Yan Chen; Zhenyu Chen; Yangming Cheng; Yutao Cui; Chunyuan Deng; Jiahua Dong; Matteo Dunnhofer; Wei Feng; Jianlong Fu; Jie Gao; Ruize Han; Zeqi Hao; Jun-Yan He; Keji He; Zhenyu He; Xiantao Hu; Kaer Huang; Yuqing Huang; Yi Jiang; Ben Kang; Jin-Peng Lan; Hyungjun Lee; Chenyang Li; Jiahao Li; Ning Li; Wangkai Li; Xiaodi Li; Xin Li; Pengyu Liu; Yue Liu; Huchuan Lu; Bin Luo; Ping Luo; Yinchao Ma; Deshui Miao; Christian Micheloni; Kannappan Palaniappan; Hancheol Park; Matthieu Paul; HouWen Peng; Zekun Qian; Gani Rahmon; Norbert Scherer-Negenborn; Pengcheng Shao; Wooksu Shin; Elham Soltani Kazemi; Tianhui Song; Rainer Stiefelhagen; Rui Sun; Chuanming Tang; Zhangyong Tang; Imad Eddine Toubal; Jack Valmadre; Joost van de Weijer; Luc Van Gool; Jash Vira; Stephane Vujasinovic; Cheng Wan; Jia Wan; Dong Wang; Fei Wang; Feifan Wang; He Wang; Limin Wang; Song Wang; Yaowei Wang; Zhepeng Wang; Gangshan Wu; Jiannan Wu; Qiangqiang Wu; Xiaojun Wu; Anqi Xiao; Jinxia Xie; Chenlong Xu; Min Xu; Tianyang Xu; Yuanyou Xu; Bin Yan; Dawei Yang; Ming-Hsuan Yang; Tianyu Yang; Yi Yang; Zongxin Yang; Xuanwu Yin; Fisher Yu; Hongyuan Yu; Qianjin Yu; Weichen Yu; YongSheng Yuan; Zehuan Yuan; Jianlin Zhang; Lu Zhang; Tianzhu Zhang; Guodongfang Zhao; Shaochuan Zhao; Yaozong Zheng; Bineng Zhong; Jiawen Zhu; Xuefeng Zhu; Yueting Zhuang; ChengAo Zong; Kunlong Zuo
	Title	The First Visual Object Tracking Segmentation VOTS2023 Challenge Results			Type	Conference Article
	Year	2023	Publication	Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops	Abbreviated Journal
	Volume		Issue		Pages	1796-1818
	Keywords
	Abstract	The Visual Object Tracking Segmentation VOTS2023 challenge is the eleventh annual tracker benchmarking activity of the VOT initiative. This challenge is the first to merge short-term and long-term as well as single-target and multiple-target tracking with segmentation masks as the only target location specification. A new dataset was created; the ground truth has been withheld to prevent overfitting. New performance measures and evaluation protocols have been created along with a new toolkit and an evaluation server. Results of the presented 47 trackers indicate that modern tracking frameworks are well-suited to deal with convergence of short-term and long-term tracking and that multiple and single target tracking can be considered a single problem. A leaderboard, with participating trackers details, the source code, the datasets, and the evaluation kit are publicly available at the challenge website\footnote https://www.votchallenge.net/vots2023/.
	Address	Paris; France; October 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCVW
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ KMD2023			Serial	3939
Permanent link to this record



	Author	Valeriya Khan; Sebastian Cygert; Bartlomiej Twardowski; Tomasz Trzcinski
	Title	Looking Through the Past: Better Knowledge Retention for Generative Replay in Continual Learning			Type	Conference Article
	Year	2023	Publication	Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops	Abbreviated Journal
	Volume		Issue		Pages	3496-3500
	Keywords
	Abstract	In this work, we improve the generative replay in a continual learning setting. We notice that in VAE-based generative replay, the generated features are quite far from the original ones when mapped to the latent space. Therefore, we propose modifications that allow the model to learn and generate complex data. More specifically, we incorporate the distillation in latent space between the current and previous models to reduce feature drift. Additionally, a latent matching for the reconstruction and original data is proposed to improve generated features alignment. Further, based on the observation that the reconstructions are better for preserving knowledge, we add the cycling of generations through the previously trained model to make them closer to the original data. Our method outperforms other generative replay methods in various scenarios.
	Address
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCVW
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ KCT2023			Serial	3942
Permanent link to this record



	Author	Damian Sojka; Sebastian Cygert; Bartlomiej Twardowski; Tomasz Trzcinski
	Title	AR-TTA: A Simple Method for Real-World Continual Test-Time Adaptation			Type	Conference Article
	Year	2023	Publication	Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops	Abbreviated Journal
	Volume		Issue		Pages	3491-3495
	Keywords
	Abstract	Test-time adaptation is a promising research direction that allows the source model to adapt itself to changes in data distribution without any supervision. Yet, current methods are usually evaluated on benchmarks that are only a simplification of real-world scenarios. Hence, we propose to validate test-time adaptation methods using the recently introduced datasets for autonomous driving, namely CLAD-C and SHIFT. We observe that current test-time adaptation methods struggle to effectively handle varying degrees of domain shift, often resulting in degraded performance that falls below that of the source model. We noticed that the root of the problem lies in the inability to preserve the knowledge of the source model and adapt to dynamically changing, temporally correlated data streams. Therefore, we enhance well-established self-training framework by incorporating a small memory buffer to increase model stability and at the same time perform dynamic adaptation based on the intensity of domain shift. The proposed method, named AR-TTA, outperforms existing approaches on both synthetic and more real-world benchmarks and shows robustness across a variety of TTA scenarios.
	Address	Paris; France; October 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCVW
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ SCT2023			Serial	3943
Permanent link to this record



	Author	Filip Szatkowski; Mateusz Pyla; Marcin Przewięzlikowski; Sebastian Cygert; Bartłomiej Twardowski; Tomasz Trzcinski
	Title	Adapt Your Teacher: Improving Knowledge Distillation for Exemplar-Free Continual Learning			Type	Conference Article
	Year	2023	Publication	Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops	Abbreviated Journal
	Volume		Issue		Pages	3512-3517
	Keywords
	Abstract	In this work, we investigate exemplar-free class incremental learning (CIL) with knowledge distillation (KD) as a regularization strategy, aiming to prevent forgetting. KD-based methods are successfully used in CIL, but they often struggle to regularize the model without access to exemplars of the training data from previous tasks. Our analysis reveals that this issue originates from substantial representation shifts in the teacher network when dealing with out-of-distribution data. This causes large errors in the KD loss component, leading to performance degradation in CIL. Inspired by recent test-time adaptation methods, we introduce Teacher Adaptation (TA), a method that concurrently updates the teacher and the main model during incremental training. Our method seamlessly integrates with KD-based CIL approaches and allows for consistent enhancement of their performance across multiple exemplar-free CIL benchmarks.
	Address	Paris; France; October 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCVW
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @			Serial	3944
Permanent link to this record



	Author	Fei Wang; Kai Wang; Joost Van de Weijer
	Title	ScrollNet: DynamicWeight Importance for Continual Learning			Type	Conference Article
	Year	2023	Publication	Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops	Abbreviated Journal
	Volume		Issue		Pages	3345-3355
	Keywords
	Abstract	The principle underlying most existing continual learning (CL) methods is to prioritize stability by penalizing changes in parameters crucial to old tasks, while allowing for plasticity in other parameters. The importance of weights for each task can be determined either explicitly through learning a task-specific mask during training (e.g., parameter isolation-based approaches) or implicitly by introducing a regularization term (e.g., regularization-based approaches). However, all these methods assume that the importance of weights for each task is unknown prior to data exposure. In this paper, we propose ScrollNet as a scrolling neural network for continual learning. ScrollNet can be seen as a dynamic network that assigns the ranking of weight importance for each task before data exposure, thus achieving a more favorable stability-plasticity tradeoff during sequential task learning by reassigning this ranking for different tasks. Additionally, we demonstrate that ScrollNet can be combined with various CL methods, including regularization-based and replay-based approaches. Experimental results on CIFAR100 and TinyImagenet datasets show the effectiveness of our proposed method.
	Address	Paris; France; October 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCVW
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ WWW2023			Serial	3945
Permanent link to this record



	Author	Dawid Rymarczyk; Joost van de Weijer; Bartosz Zielinski; Bartlomiej Twardowski
	Title	ICICLE: Interpretable Class Incremental Continual Learning. Dawid Rymarczyk			Type	Conference Article
	Year	2023	Publication	20th IEEE International Conference on Computer Vision	Abbreviated Journal
	Volume		Issue		Pages	1887-1898
	Keywords
	Abstract	Continual learning enables incremental learning of new tasks without forgetting those previously learned, resulting in positive knowledge transfer that can enhance performance on both new and old tasks. However, continual learning poses new challenges for interpretability, as the rationale behind model predictions may change over time, leading to interpretability concept drift. We address this problem by proposing Interpretable Class-InCremental LEarning (ICICLE), an exemplar-free approach that adopts a prototypical part-based approach. It consists of three crucial novelties: interpretability regularization that distills previously learned concepts while preserving user-friendly positive reasoning; proximity-based prototype initialization strategy dedicated to the fine-grained setting; and task-recency bias compensation devoted to prototypical parts. Our experimental results demonstrate that ICICLE reduces the interpretability concept drift and outperforms the existing exemplar-free methods of common class-incremental learning when applied to concept-based models.
	Address	Paris; France; October 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCV
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ RWZ2023			Serial	3947
Permanent link to this record



	Author	Yuyang Liu; Yang Cong; Dipam Goswami; Xialei Liu; Joost Van de Weijer
	Title	Augmented Box Replay: Overcoming Foreground Shift for Incremental Object Detection			Type	Conference Article
	Year	2023	Publication	20th IEEE International Conference on Computer Vision	Abbreviated Journal
	Volume		Issue		Pages	11367-11377
	Keywords
	Abstract	In incremental learning, replaying stored samples from previous tasks together with current task samples is one of the most efficient approaches to address catastrophic forgetting. However, unlike incremental classification, image replay has not been successfully applied to incremental object detection (IOD). In this paper, we identify the overlooked problem of foreground shift as the main reason for this. Foreground shift only occurs when replaying images of previous tasks and refers to the fact that their background might contain foreground objects of the current task. To overcome this problem, a novel and efficient Augmented Box Replay (ABR) method is developed that only stores and replays foreground objects and thereby circumvents the foreground shift problem. In addition, we propose an innovative Attentive RoI Distillation loss that uses spatial attention from region-of-interest (RoI) features to constrain current model to focus on the most important information from old model. ABR significantly reduces forgetting of previous classes while maintaining high plasticity in current classes. Moreover, it considerably reduces the storage requirements when compared to standard image replay. Comprehensive experiments on Pascal-VOC and COCO datasets support the state-of-the-art performance of our model.
	Address	Paris; France; October 2023
	Corporate Author				Thesis
	Publisher		Place of Publication		Editor
	Language		Summary Language		Original Title
	Series Editor		Series Title		Abbreviated Series Title
	Series Volume		Series Issue		Edition
	ISSN		ISBN		Medium
	Area		Expedition		Conference	ICCV
	Notes	LAMP			Approved	no
	Call Number	Admin @ si @ LCG2023			Serial	3949
Permanent link to this record