toggle visibility Search & Display Options

Select All    Deselect All
 |   | 
Details
   print
  Records Links
Author Francesco Pelosin; Saurav Jha; Andrea Torsello; Bogdan Raducanu; Joost Van de Weijer edit   pdf
url  doi
openurl 
  Title Towards exemplar-free continual learning in vision transformers: an account of attention, functional and weight regularization Type Conference Article
  Year 2022 Publication IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) Abbreviated Journal  
  Volume Issue Pages  
  Keywords Learning systems; Weight measurement; Image recognition; Surgery; Benchmark testing; Transformers; Stability analysis  
  Abstract In this paper, we investigate the continual learning of Vision Transformers (ViT) for the challenging exemplar-free scenario, with special focus on how to efficiently distill the knowledge of its crucial self-attention mechanism (SAM). Our work takes an initial step towards a surgical investigation of SAM for designing coherent continual learning methods in ViTs. We first carry out an evaluation of established continual learning regularization techniques. We then examine the effect of regularization when applied to two key enablers of SAM: (a) the contextualized embedding layers, for their ability to capture well-scaled representations with respect to the values, and (b) the prescaled attention maps, for carrying value-independent global contextual information. We depict the perks of each distilling strategy on two image recognition benchmarks (CIFAR100 and ImageNet-32) – while (a) leads to a better overall accuracy, (b) helps enhance the rigidity by maintaining competitive performances. Furthermore, we identify the limitation imposed by the symmetric nature of regularization losses. To alleviate this, we propose an asymmetric variant and apply it to the pooled output distillation (POD) loss adapted for ViTs. Our experiments confirm that introducing asymmetry to POD boosts its plasticity while retaining stability across (a) and (b). Moreover, we acknowledge low forgetting measures for all the compared methods, indicating that ViTs might be naturally inclined continual learners. 1  
  Address (down) New Orleans; USA; June 2022  
  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 CVPRW  
  Notes LAMP; 600.147 Approved no  
  Call Number Admin @ si @ PJT2022 Serial 3784  
Permanent link to this record
 

 
Author Hector Laria Mantecon; Yaxing Wang; Joost Van de Weijer; Bogdan Raducanu edit  openurl
  Title Transferring Unconditional to Conditional GANs With Hyper-Modulation Type Conference Article
  Year 2022 Publication IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) Abbreviated Journal  
  Volume Issue Pages  
  Keywords  
  Abstract GANs have matured in recent years and are able to generate high-resolution, realistic images. However, the computational resources and the data required for the training of high-quality GANs are enormous, and the study of transfer learning of these models is therefore an urgent topic. Many of the available high-quality pretrained GANs are unconditional (like StyleGAN). For many applications, however, conditional GANs are preferable, because they provide more control over the generation process, despite often suffering more training difficulties. Therefore, in this paper, we focus on transferring from high-quality pretrained unconditional GANs to conditional GANs. This requires architectural adaptation of the pretrained GAN to perform the conditioning. To this end, we propose hyper-modulated generative networks that allow for shared and complementary supervision. To prevent the additional weights of the hypernetwork to overfit, with subsequent mode collapse on small target domains, we introduce a self-initialization procedure that does not require any real data to initialize the hypernetwork parameters. To further improve the sample efficiency of the transfer, we apply contrastive learning in the discriminator, which effectively works on very limited batch sizes. In extensive experiments, we validate the efficiency of the hypernetworks, self-initialization and contrastive loss for knowledge transfer on standard benchmarks.  
  Address (down) New Orleans; USA; June 2022  
  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 CVPRW  
  Notes LAMP; 600.147; 602.200 Approved no  
  Call Number LWW2022a Serial 3785  
Permanent link to this record
 

 
Author Dipam Goswami; Yuyang Liu ; Bartlomiej Twardowski; Joost Van de Weijer edit  url
openurl 
  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 (down) 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 edit  url
openurl 
  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 (down) 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 edit   pdf
url  openurl
  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 (down) 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 Kai Wang; Xialei Liu; Andrew Bagdanov; Luis Herranz; Shangling Jui; Joost Van de Weijer edit   pdf
doi  openurl
  Title Incremental Meta-Learning via Episodic Replay Distillation for Few-Shot Image Recognition Type Conference Article
  Year 2022 Publication CVPR 2022 Workshop on Continual Learning (CLVision, 3rd Edition) Abbreviated Journal  
  Volume Issue Pages 3728-3738  
  Keywords Training; Computer vision; Image recognition; Upper bound; Conferences; Pattern recognition; Task analysis  
  Abstract In this paper we consider the problem of incremental meta-learning in which classes are presented incrementally in discrete tasks. We propose Episodic Replay Distillation (ERD), that mixes classes from the current task with exemplars from previous tasks when sampling episodes for meta-learning. To allow the training to benefit from a large as possible variety of classes, which leads to more gener-
alizable feature representations, we propose the cross-task meta loss. Furthermore, we propose episodic replay distillation that also exploits exemplars for improved knowledge distillation. Experiments on four datasets demonstrate that ERD surpasses the state-of-the-art. In particular, on the more challenging one-shot, long task sequence scenarios, we reduce the gap between Incremental Meta-Learning and
the joint-training upper bound from 3.5% / 10.1% / 13.4% / 11.7% with the current state-of-the-art to 2.6% / 2.9% / 5.0% / 0.2% with our method on Tiered-ImageNet / Mini-ImageNet / CIFAR100 / CUB, respectively.
 
  Address (down) New Orleans, USA; 20 June 2022  
  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 CVPRW  
  Notes LAMP; 600.147 Approved no  
  Call Number Admin @ si @ WLB2022 Serial 3686  
Permanent link to this record
 

 
Author Bojana Gajic; Ariel Amato; Ramon Baldrich; Joost Van de Weijer; Carlo Gatta edit   pdf
doi  openurl
  Title Area Under the ROC Curve Maximization for Metric Learning Type Conference Article
  Year 2022 Publication CVPR 2022 Workshop on Efficien Deep Learning for Computer Vision (ECV 2022, 5th Edition) Abbreviated Journal  
  Volume Issue Pages  
  Keywords Training; Computer vision; Conferences; Area measurement; Benchmark testing; Pattern recognition  
  Abstract Most popular metric learning losses have no direct relation with the evaluation metrics that are subsequently applied to evaluate their performance. We hypothesize that training a metric learning model by maximizing the area under the ROC curve (which is a typical performance measure of recognition systems) can induce an implicit ranking suitable for retrieval problems. This hypothesis is supported by previous work that proved that a curve dominates in ROC space if and only if it dominates in Precision-Recall space. To test this hypothesis, we design and maximize an approximated, derivable relaxation of the area under the ROC curve. The proposed AUC loss achieves state-of-the-art results on two large scale retrieval benchmark datasets (Stanford Online Products and DeepFashion In-Shop). Moreover, the AUC loss achieves comparable performance to more complex, domain specific, state-of-the-art methods for vehicle re-identification.  
  Address (down) New Orleans, USA; 20 June 2022  
  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 CVPRW  
  Notes CIC; LAMP; Approved no  
  Call Number Admin @ si @ GAB2022 Serial 3700  
Permanent link to this record
 

 
Author Alex Gomez-Villa; Bartlomiej Twardowski; Lu Yu; Andrew Bagdanov; Joost Van de Weijer edit   pdf
doi  openurl
  Title Continually Learning Self-Supervised Representations With Projected Functional Regularization Type Conference Article
  Year 2022 Publication CVPR 2022 Workshop on Continual Learning (CLVision, 3rd Edition) Abbreviated Journal  
  Volume Issue Pages 3866-3876  
  Keywords Computer vision; Conferences; Self-supervised learning; Image representation; Pattern recognition  
  Abstract Recent self-supervised learning methods are able to learn high-quality image representations and are closing the gap with supervised approaches. However, these methods are unable to acquire new knowledge incrementally – they are, in fact, mostly used only as a pre-training phase over IID data. In this work we investigate self-supervised methods in continual learning regimes without any replay
mechanism. We show that naive functional regularization,also known as feature distillation, leads to lower plasticity and limits continual learning performance. Instead, we propose Projected Functional Regularization in which a separate temporal projection network ensures that the newly learned feature space preserves information of the previous one, while at the same time allowing for the learning of new features. This prevents forgetting while maintaining the plasticity of the learner. Comparison with other incremental learning approaches applied to self-supervision demonstrates that our method obtains competitive performance in
different scenarios and on multiple datasets.
 
  Address (down) New Orleans, USA; 20 June 2022  
  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 CVPRW  
  Notes LAMP: 600.147; 600.120 Approved no  
  Call Number Admin @ si @ GTY2022 Serial 3704  
Permanent link to this record
 

 
Author Mohamed Ramzy Ibrahim; Robert Benavente; Felipe Lumbreras; Daniel Ponsa edit   pdf
doi  openurl
  Title 3DRRDB: Super Resolution of Multiple Remote Sensing Images using 3D Residual in Residual Dense Blocks Type Conference Article
  Year 2022 Publication CVPR 2022 Workshop on IEEE Perception Beyond the Visible Spectrum workshop series (PBVS, 18th Edition) Abbreviated Journal  
  Volume Issue Pages  
  Keywords Training; Solid modeling; Three-dimensional displays; PSNR; Convolution; Superresolution; Pattern recognition  
  Abstract The rapid advancement of Deep Convolutional Neural Networks helped in solving many remote sensing problems, especially the problems of super-resolution. However, most state-of-the-art methods focus more on Single Image Super-Resolution neglecting Multi-Image Super-Resolution. In this work, a new proposed 3D Residual in Residual Dense Blocks model (3DRRDB) focuses on remote sensing Multi-Image Super-Resolution for two different single spectral bands. The proposed 3DRRDB model explores the idea of 3D convolution layers in deeply connected Dense Blocks and the effect of local and global residual connections with residual scaling in Multi-Image Super-Resolution. The model tested on the Proba-V challenge dataset shows a significant improvement above the current state-of-the-art models scoring a Corrected Peak Signal to Noise Ratio (cPSNR) of 48.79 dB and 50.83 dB for Near Infrared (NIR) and RED Bands respectively. Moreover, the proposed 3DRRDB model scores a Corrected Structural Similarity Index Measure (cSSIM) of 0.9865 and 0.9909 for NIR and RED bands respectively.  
  Address (down) New Orleans, USA; 19 June 2022  
  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 CVPRW  
  Notes MSIAU; 600.130 Approved no  
  Call Number Admin @ si @ IBL2022 Serial 3693  
Permanent link to this record
 

 
Author Angel Sappa; M.A. Garcia edit  openurl
  Title Hierarchical Clustering of 3D Objects and its Application to Minimum Distance Computation Type Conference Article
  Year 2004 Publication IEEE International Conference on Robotics & Automation, 5287–5292, New Orleans, LA (USA), ISBN: 0–7803–8232–3 Abbreviated Journal  
  Volume Issue Pages  
  Keywords  
  Abstract  
  Address (down) New Orleans, LA, USA  
  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  
  Notes ADAS Approved no  
  Call Number ADAS @ adas @ SaG2004b Serial 459  
Permanent link to this record
 

 
Author N. Zakaria; Jean-Marc Ogier; Josep Llados edit  openurl
  Title The Fuzzy-Spatial Descriptor for the Online Graphic Recognition: Overlapping Matrix Algorithm Type Book Chapter
  Year 2006 Publication 7th International Workshop, Document Analysis Systems VII (DAS´06), LNCS 3872: 616–627 Abbreviated Journal  
  Volume Issue Pages  
  Keywords  
  Abstract  
  Address (down) Nelson (New Zealand)  
  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  
  Notes DAG Approved no  
  Call Number DAG @ dag @ ZOL2006 Serial 629  
Permanent link to this record
 

 
Author T.O. Nguyen; Salvatore Tabbone; Oriol Ramos Terrades edit  openurl
  Title Symbol Descriptor Based on Shape Context and Vector Model of Information Retrieval Type Conference Article
  Year 2008 Publication Proceedings of the 8th IAPR International Workshop on Document Analysis Systems, Abbreviated Journal  
  Volume Issue Pages 191-197  
  Keywords  
  Abstract  
  Address (down) Nara, Japan  
  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 DAS  
  Notes DAG Approved no  
  Call Number Admin @ si @ NTR2008a Serial 1873  
Permanent link to this record
 

 
Author Partha Pratim Roy; Umapada Pal; Josep Llados edit  openurl
  Title Multi-oriented English Text Line Extraction using Background and Foreground Information Type Conference Article
  Year 2008 Publication Proceedings of the 8th IAPR International Workshop on Document Analysis Systems, Abbreviated Journal  
  Volume Issue Pages 315–322  
  Keywords  
  Abstract  
  Address (down) Nara (Japo)  
  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 DAS  
  Notes DAG Approved no  
  Call Number DAG @ dag @ RPL2008b Serial 1047  
Permanent link to this record
 

 
Author Marçal Rusiñol; Josep Llados edit  openurl
  Title Word and Symbol Spotting using Spatial Organization of Local Descriptors Type Conference Article
  Year 2008 Publication Proceedings of the 8th IAPR International Workshop on Document Analysis Systems, Abbreviated Journal  
  Volume Issue Pages 489–496  
  Keywords  
  Abstract  
  Address (down) Nara (Japan)  
  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 DAS  
  Notes DAG Approved no  
  Call Number DAG @ dag @ RuL2008b Serial 1059  
Permanent link to this record
 

 
Author Mathieu Nicolas Delalandre; Ernest Valveny; Josep Llados edit  openurl
  Title Performance Evaluation of Symbol Recognition and Spotting Systems Type Conference Article
  Year 2008 Publication Proceedings of the 8th International Workshop on Document Analysis Systems, Abbreviated Journal  
  Volume Issue Pages 497–505  
  Keywords  
  Abstract  
  Address (down) Nara (Japan)  
  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 DAS  
  Notes DAG Approved no  
  Call Number DAG @ dag @ DVL2008b Serial 1060  
Permanent link to this record
Select All    Deselect All
 |   | 
Details
   print

Save Citations:
Export Records: