Search Results for author: Tinne Tuytelaars

Found 136 papers, 55 papers with code

More Classifiers, Less Forgetting: A Generic Multi-classifier Paradigm for Incremental Learning

1 code implementation ECCV 2020 Yu Liu, Sarah Parisot, Gregory Slabaugh, Xu Jia, Ales Leonardis, Tinne Tuytelaars

Since those regularization strategies are mostly associated with classifier outputs, we propose a MUlti-Classifier (MUC) incremental learning paradigm that integrates an ensemble of auxiliary classifiers to estimate more effective regularization constraints.

Incremental Learning

Animate Your Motion: Turning Still Images into Dynamic Videos

no code implementations15 Mar 2024 Mingxiao Li, Bo Wan, Marie-Francine Moens, Tinne Tuytelaars

For the first time, we integrate both semantic and motion cues within a diffusion model for video generation, as demonstrated in Fig 1.

Specificity Text-to-Video Generation +1

Introducing Routing Functions to Vision-Language Parameter-Efficient Fine-Tuning with Low-Rank Bottlenecks

no code implementations14 Mar 2024 Tingyu Qu, Tinne Tuytelaars, Marie-Francine Moens

Also when fine-tuning a pre-trained multimodal model such as CLIP-BART, we observe smaller but consistent improvements across a range of VL PEFT tasks.

The Common Stability Mechanism behind most Self-Supervised Learning Approaches

1 code implementation22 Feb 2024 Abhishek Jha, Matthew B. Blaschko, Yuki M. Asano, Tinne Tuytelaars

Last couple of years have witnessed a tremendous progress in self-supervised learning (SSL), the success of which can be attributed to the introduction of useful inductive biases in the learning process to learn meaningful visual representations while avoiding collapse.

Self-Supervised Learning

Infinite dSprites for Disentangled Continual Learning: Separating Memory Edits from Generalization

no code implementations27 Dec 2023 Sebastian Dziadzio, Çağatay Yıldız, Gido M. van de Ven, Tomasz Trzciński, Tinne Tuytelaars, Matthias Bethge

In a simple setting with direct supervision on the generative factors, we show how learning class-agnostic transformations offers a way to circumvent catastrophic forgetting and improve classification accuracy over time.

Classification Continual Learning +3

Estimating calibration error under label shift without labels

no code implementations14 Dec 2023 Teodora Popordanoska, Gorjan Radevski, Tinne Tuytelaars, Matthew B. Blaschko

In the face of dataset shift, model calibration plays a pivotal role in ensuring the reliability of machine learning systems.

TeTriRF: Temporal Tri-Plane Radiance Fields for Efficient Free-Viewpoint Video

no code implementations10 Dec 2023 Minye Wu, Zehao Wang, Georgios Kouros, Tinne Tuytelaars

Neural Radiance Fields (NeRF) revolutionize the realm of visual media by providing photorealistic Free-Viewpoint Video (FVV) experiences, offering viewers unparalleled immersion and interactivity.

NeVRF: Neural Video-based Radiance Fields for Long-duration Sequences

no code implementations10 Dec 2023 Minye Wu, Tinne Tuytelaars

Our extensive experiments demonstrate the effectiveness of NeVRF in enabling long-duration sequence rendering, sequential data reconstruction, and compact data storage.

Continual Learning Novel View Synthesis

Continual Learning of Diffusion Models with Generative Distillation

1 code implementation23 Nov 2023 Sergi Masip, Pau Rodriguez, Tinne Tuytelaars, Gido M. van de Ven

We demonstrate that our approach significantly improves the continual learning performance of generative replay with only a moderate increase in the computational costs.

Continual Learning Denoising +1

Contrastive Learning for Multi-Object Tracking with Transformers

no code implementations14 Nov 2023 Pierre-François De Plaen, Nicola Marinello, Marc Proesmans, Tinne Tuytelaars, Luc van Gool

The DEtection TRansformer (DETR) opened new possibilities for object detection by modeling it as a translation task: converting image features into object-level representations.

Contrastive Learning Multi-Object Tracking +4

Two Complementary Perspectives to Continual Learning: Ask Not Only What to Optimize, But Also How

no code implementations8 Nov 2023 Timm Hess, Tinne Tuytelaars, Gido M. van de Ven

Recent years have seen considerable progress in the continual training of deep neural networks, predominantly thanks to approaches that add replay or regularization terms to the loss function to approximate the joint loss over all tasks so far.

Continual Learning

CrIBo: Self-Supervised Learning via Cross-Image Object-Level Bootstrapping

1 code implementation11 Oct 2023 Tim Lebailly, Thomas Stegmüller, Behzad Bozorgtabar, Jean-Philippe Thiran, Tinne Tuytelaars

Leveraging nearest neighbor retrieval for self-supervised representation learning has proven beneficial with object-centric images.

In-Context Learning Object +3

Exploiting CLIP for Zero-shot HOI Detection Requires Knowledge Distillation at Multiple Levels

1 code implementation10 Sep 2023 Bo Wan, Tinne Tuytelaars

In this paper, we investigate the task of zero-shot human-object interaction (HOI) detection, a novel paradigm for identifying HOIs without the need for task-specific annotations.

Human-Object Interaction Detection Knowledge Distillation +1

Ref-DVGO: Reflection-Aware Direct Voxel Grid Optimization for an Improved Quality-Efficiency Trade-Off in Reflective Scene Reconstruction

1 code implementation16 Aug 2023 Georgios Kouros, Minye Wu, Shubham Shrivastava, Sushruth Nagesh, Punarjay Chakravarty, Tinne Tuytelaars

To this end, we investigate an implicit-explicit approach based on conventional volume rendering to enhance the reconstruction quality and accelerate the training and rendering processes.

Novel View Synthesis

Visually-Aware Context Modeling for News Image Captioning

no code implementations16 Aug 2023 Tingyu Qu, Tinne Tuytelaars, Marie-Francine Moens

On top of that, humans often play a central role in news stories, as also proven by the face-name co-occurrence pattern we discover in existing News Image Captioning datasets.

Image Captioning

Multimodal Distillation for Egocentric Action Recognition

1 code implementation ICCV 2023 Gorjan Radevski, Dusan Grujicic, Marie-Francine Moens, Matthew Blaschko, Tinne Tuytelaars

The goal of this work is to retain the performance of such a multimodal approach, while using only the RGB frames as input at inference time.

Action Recognition Knowledge Distillation +2

EffSeg: Efficient Fine-Grained Instance Segmentation using Structure-Preserving Sparsity

1 code implementation4 Jul 2023 Cédric Picron, Tinne Tuytelaars

In this work, we propose EffSeg performing fine-grained instance segmentation in an efficient way by using our Structure-Preserving Sparsity (SPS) method based on separately storing the active features, the passive features and a dense 2D index map containing the feature indices.

Instance Segmentation Segmentation +1

Continual Learning with Pretrained Backbones by Tuning in the Input Space

no code implementations5 Jun 2023 Simone Marullo, Matteo Tiezzi, Marco Gori, Stefano Melacci, Tinne Tuytelaars

The intrinsic difficulty in adapting deep learning models to non-stationary environments limits the applicability of neural networks to real-world tasks.

Continual Learning Image Classification

Few-Shot Open-Set Learning for On-Device Customization of KeyWord Spotting Systems

1 code implementation3 Jun 2023 Manuele Rusci, Tinne Tuytelaars

A personalized KeyWord Spotting (KWS) pipeline typically requires the training of a Deep Learning model on a large set of user-defined speech utterances, preventing fast customization directly applied on-device.

Few-Shot Learning Keyword Spotting +1

Prediction Error-based Classification for Class-Incremental Learning

1 code implementation30 May 2023 Michał Zając, Tinne Tuytelaars, Gido M. van de Ven

Class-incremental learning (CIL) is a particularly challenging variant of continual learning, where the goal is to learn to discriminate between all classes presented in an incremental fashion.

Classification Class Incremental Learning +1

Neural Residual Radiance Fields for Streamably Free-Viewpoint Videos

no code implementations CVPR 2023 Liao Wang, Qiang Hu, Qihan He, Ziyu Wang, Jingyi Yu, Tinne Tuytelaars, Lan Xu, Minye Wu

The success of the Neural Radiance Fields (NeRFs) for modeling and free-view rendering static objects has inspired numerous attempts on dynamic scenes.

Neural Rendering

Knowledge Accumulation in Continually Learned Representations and the Issue of Feature Forgetting

no code implementations3 Apr 2023 Timm Hess, Eli Verwimp, Gido M. van de Ven, Tinne Tuytelaars

Carefully taking both aspects into account, we show that, even though it is true that feature forgetting can be small in absolute terms, newly learned information tends to be forgotten just as catastrophically at the level of the representation as it is at the output level.

Continual Learning Image Classification +2

Adaptive Similarity Bootstrapping for Self-Distillation based Representation Learning

1 code implementation ICCV 2023 Tim Lebailly, Thomas Stegmüller, Behzad Bozorgtabar, Jean-Philippe Thiran, Tinne Tuytelaars

Most self-supervised methods for representation learning leverage a cross-view consistency objective i. e., they maximize the representation similarity of a given image's augmented views.

Contrastive Learning Representation Learning

CrOC: Cross-View Online Clustering for Dense Visual Representation Learning

2 code implementations CVPR 2023 Thomas Stegmüller, Tim Lebailly, Behzad Bozorgtabar, Tinne Tuytelaars, Jean-Philippe Thiran

More importantly, the clustering algorithm conjointly operates on the features of both views, thereby elegantly bypassing the issue of content not represented in both views and the ambiguous matching of objects from one crop to the other.

Clustering Online Clustering +5

Weakly-supervised HOI Detection via Prior-guided Bi-level Representation Learning

no code implementations2 Mar 2023 Bo Wan, Yongfei Liu, Desen Zhou, Tinne Tuytelaars, Xuming He

Human object interaction (HOI) detection plays a crucial role in human-centric scene understanding and serves as a fundamental building-block for many vision tasks.

Human-Object Interaction Detection Knowledge Distillation +3

Layout-aware Dreamer for Embodied Referring Expression Grounding

1 code implementation30 Nov 2022 Mingxiao Li, Zehao Wang, Tinne Tuytelaars, Marie-Francine Moens

In this work, we study the problem of Embodied Referring Expression Grounding, where an agent needs to navigate in a previously unseen environment and localize a remote object described by a concise high-level natural language instruction.

Common Sense Reasoning Navigate +1

Weakly Supervised Face Naming with Symmetry-Enhanced Contrastive Loss

no code implementations17 Oct 2022 Tingyu Qu, Tinne Tuytelaars, Marie-Francine Moens

We revisit the weakly supervised cross-modal face-name alignment task; that is, given an image and a caption, we label the faces in the image with the names occurring in the caption.

Contrastive Learning

Students taught by multimodal teachers are superior action recognizers

no code implementations9 Oct 2022 Gorjan Radevski, Dusan Grujicic, Matthew Blaschko, Marie-Francine Moens, Tinne Tuytelaars

Our approach is based on multimodal knowledge distillation, featuring a multimodal teacher (in the current experiments trained only using object detections, optical flow and RGB frames) and a unimodal student (using only RGB frames as input).

Action Recognition Knowledge Distillation +3

CLAD: A realistic Continual Learning benchmark for Autonomous Driving

1 code implementation7 Oct 2022 Eli Verwimp, Kuo Yang, Sarah Parisot, Hong Lanqing, Steven McDonagh, Eduardo Pérez-Pellitero, Matthias De Lange, Tinne Tuytelaars

In this paper we describe the design and the ideas motivating a new Continual Learning benchmark for Autonomous Driving (CLAD), that focuses on the problems of object classification and object detection.

Autonomous Driving Continual Learning +3

FQDet: Fast-converging Query-based Detector

2 code implementations5 Oct 2022 Cédric Picron, Punarjay Chakravarty, Tinne Tuytelaars

Recently, two-stage Deformable DETR introduced the query-based two-stage head, a new type of two-stage head different from the region-based two-stage heads of classical detectors as Faster R-CNN.

Global-Local Self-Distillation for Visual Representation Learning

1 code implementation29 Jul 2022 Tim Lebailly, Tinne Tuytelaars

The downstream accuracy of self-supervised methods is tightly linked to the proxy task solved during training and the quality of the gradients extracted from it.

Representation Learning

Continual evaluation for lifelong learning: Identifying the stability gap

1 code implementation26 May 2022 Matthias De Lange, Gido van de Ven, Tinne Tuytelaars

Despite the progress in the field of continual learning to overcome this forgetting, we show that a set of common state-of-the-art methods still suffers from substantial forgetting upon starting to learn new tasks, except that this forgetting is temporary and followed by a phase of performance recovery.

Continual Learning Incremental Learning +1

Continual Pre-Training Mitigates Forgetting in Language and Vision

1 code implementation19 May 2022 Andrea Cossu, Tinne Tuytelaars, Antonio Carta, Lucia Passaro, Vincenzo Lomonaco, Davide Bacciu

We formalize and investigate the characteristics of the continual pre-training scenario in both language and vision environments, where a model is continually pre-trained on a stream of incoming data and only later fine-tuned to different downstream tasks.

Continual Learning Continual Pretraining

Spatial Consistency Loss for Training Multi-Label Classifiers from Single-Label Annotations

no code implementations11 Mar 2022 Thomas Verelst, Paul K. Rubenstein, Marcin Eichner, Tinne Tuytelaars, Maxim Berman

We show that adding a consistency loss, ensuring that the predictions of the network are consistent over consecutive training epochs, is a simple yet effective method to train multi-label classifiers in a weakly supervised setting.

Data Augmentation Multi-Label Classification +1

New Insights on Reducing Abrupt Representation Change in Online Continual Learning

3 code implementations ICLR 2022 Lucas Caccia, Rahaf Aljundi, Nader Asadi, Tinne Tuytelaars, Joelle Pineau, Eugene Belilovsky

In this work, we focus on the change in representations of observed data that arises when previously unobserved classes appear in the incoming data stream, and new classes must be distinguished from previous ones.

Class Incremental Learning

Barlow constrained optimization for Visual Question Answering

1 code implementation7 Mar 2022 Abhishek Jha, Badri N. Patro, Luc van Gool, Tinne Tuytelaars

In this paper, we propose a novel regularization for VQA models, Constrained Optimization using Barlow's theory (COB), that improves the information content of the joint space by minimizing the redundancy.

Question Answering Visual Question Answering

Find a Way Forward: a Language-Guided Semantic Map Navigator

no code implementations7 Mar 2022 Zehao Wang, Mingxiao Li, Minye Wu, Marie-Francine Moens, Tinne Tuytelaars

In this paper, we introduce the map-language navigation task where an agent executes natural language instructions and moves to the target position based only on a given 3D semantic map.

Imitation Learning

RARA: Zero-shot Sim2Real Visual Navigation with Following Foreground Cues

1 code implementation8 Jan 2022 Klaas Kelchtermans, Tinne Tuytelaars

In this work, we tackle this gap for the specific case of camera-based navigation, formulating it as following a visual cue in the foreground with arbitrary backgrounds.

Visual Navigation

Revisiting spatio-temporal layouts for compositional action recognition

1 code implementation2 Nov 2021 Gorjan Radevski, Marie-Francine Moens, Tinne Tuytelaars

Recognizing human actions is fundamentally a spatio-temporal reasoning problem, and should be, at least to some extent, invariant to the appearance of the human and the objects involved.

Action Classification Action Detection +3

Trident Pyramid Networks: The importance of processing at the feature pyramid level for better object detection

1 code implementation8 Oct 2021 Cédric Picron, Tinne Tuytelaars

Given their importance, a computer vision network can be divided into three parts: a backbone (generating a feature pyramid), a neck (refining the feature pyramid) and a head (generating the final output).

Object object-detection +1

Towards Human-Understandable Visual Explanations: Human Imperceptible Cues Can Better Be Removed

no code implementations29 Sep 2021 Kaili Wang, Jose Oramas, Tinne Tuytelaars

Explainable AI (XAI) methods focus on explaining what a neural network has learned - in other words, identifying the features that are the most influential to the prediction.

Explainable Artificial Intelligence (XAI)

Glimpse-Attend-and-Explore: Self-Attention for Active Visual Exploration

1 code implementation ICCV 2021 Soroush Seifi, Abhishek Jha, Tinne Tuytelaars

In this paper, we propose the Glimpse-Attend-and-Explore model which: (a) employs self-attention to guide the visual exploration instead of task-specific uncertainty maps; (b) can be used for both dense and sparse prediction tasks; and (c) uses a contrastive stream to further improve the representations learned.

BlockCopy: High-Resolution Video Processing with Block-Sparse Feature Propagation and Online Policies

1 code implementation ICCV 2021 Thomas Verelst, Tinne Tuytelaars

In this paper we propose BlockCopy, a scheme that accelerates pretrained frame-based CNNs to process video more efficiently, compared to standard frame-by-frame processing.

Instance Segmentation Pedestrian Detection +2

Towards Human-Understandable Visual Explanations:Imperceptible High-frequency Cues Can Better Be Removed

no code implementations16 Apr 2021 Kaili Wang, Jose Oramas, Tinne Tuytelaars

Explainable AI (XAI) methods focus on explaining what a neural network has learned - in other words, identifying the features that are the most influential to the prediction.

Explainable Artificial Intelligence (XAI)

Rehearsal revealed: The limits and merits of revisiting samples in continual learning

1 code implementation ICCV 2021 Eli Verwimp, Matthias De Lange, Tinne Tuytelaars

Learning from non-stationary data streams and overcoming catastrophic forgetting still poses a serious challenge for machine learning research.

Continual Learning

New Insights on Reducing Abrupt Representation Change in Online Continual Learning

3 code implementations11 Apr 2021 Lucas Caccia, Rahaf Aljundi, Nader Asadi, Tinne Tuytelaars, Joelle Pineau, Eugene Belilovsky

In this work, we focus on the change in representations of observed data that arises when previously unobserved classes appear in the incoming data stream, and new classes must be distinguished from previous ones.

Continual Learning Metric Learning

SegBlocks: Block-Based Dynamic Resolution Networks for Real-Time Segmentation

1 code implementation24 Nov 2020 Thomas Verelst, Tinne Tuytelaars

For instance, our method reduces the number of floating-point operations of SwiftNet-RN18 by 60% and increases the inference speed by 50%, with only 0. 3% decrease in mIoU accuracy on Cityscapes.

Real-Time Semantic Segmentation

Decoding Language Spatial Relations to 2D Spatial Arrangements

1 code implementation Findings of the Association for Computational Linguistics 2020 Gorjan Radevski, Guillem Collell, Marie-Francine Moens, Tinne Tuytelaars

We address the problem of multimodal spatial understanding by decoding a set of language-expressed spatial relations to a set of 2D spatial arrangements in a multi-object and multi-relationship setting.

Learning to ground medical text in a 3D human atlas

1 code implementation CONLL 2020 Dusan Grujicic, Gorjan Radevski, Tinne Tuytelaars, Matthew Blaschko

In this paper, we develop a method for grounding medical text into a physically meaningful and interpretable space corresponding to a human atlas.

Phrase Grounding Visual Grounding

On the Exploration of Incremental Learning for Fine-grained Image Retrieval

1 code implementation15 Oct 2020 Wei Chen, Yu Liu, Weiping Wang, Tinne Tuytelaars, Erwin M. Bakker, Michael Lew

On the other hand, fine-tuning the learned representation only with the new classes leads to catastrophic forgetting.

Image Retrieval Incremental Learning +1

Self-Supervised Ranking for Representation Learning

no code implementations14 Oct 2020 Ali Varamesh, Ali Diba, Tinne Tuytelaars, Luc van Gool

We present a new framework for self-supervised representation learning by formulating it as a ranking problem in an image retrieval context on a large number of random views (augmentations) obtained from images.

Clustering Contrastive Learning +5

Feed-Forward On-Edge Fine-tuning Using Static Synthetic Gradient Modules

no code implementations21 Sep 2020 Robby Neven, Marian Verhelst, Tinne Tuytelaars, Toon Goedemé

By first training the SGMs in a meta-learning manner on a set of common objects, during fine-tuning, the SGMs provided the model with accurate gradients to successfully learn to grasp new objects.

Meta-Learning

Multiple Exemplars-based Hallucinationfor Face Super-resolution and Editing

no code implementations16 Sep 2020 Kaili Wang, Jose Oramas, Tinne Tuytelaars

Given a really low-resolution input image of a face (say 16x16 or 8x8 pixels), the goal of this paper is to reconstruct a high-resolution version thereof.

Super-Resolution

What My Motion tells me about Your Pose: A Self-Supervised Monocular 3D Vehicle Detector

no code implementations29 Jul 2020 Cédric Picron, Punarjay Chakravarty, Tom Roussel, Tinne Tuytelaars

We subsequently demonstrate an optimization-based monocular 3D bounding box detector built on top of the self-supervised vehicle orientation estimator without the requirement of expensive labeled data.

Autonomous Vehicles Domain Adaptation +1

Attend and Segment: Attention Guided Active Semantic Segmentation

no code implementations ECCV 2020 Soroush Seifi, Tinne Tuytelaars

The main idea is to refine an agent's understanding of the environment by attending the areas it is most uncertain about.

Semantic Segmentation

MIX'EM: Unsupervised Image Classification using a Mixture of Embeddings

no code implementations18 Jul 2020 Ali Varamesh, Tinne Tuytelaars

We introduce three techniques to successfully train MIX'EM and avoid degenerate solutions; (i) diversify mixture components by maximizing entropy, (ii) minimize instance conditioned component entropy to enforce a clustered embedding space, and (iii) use an associative embedding loss to enforce semantic separability.

Classification Clustering +3

Self-supervised context-aware COVID-19 document exploration through atlas grounding

1 code implementation ACL 2020 Dusan Grujicic, Gorjan Radevski, Tinne Tuytelaars, Matthew Blaschko

In this paper, we aim to develop a self-supervised grounding of Covid-related medical text based on the actual spatial relationships between the referred anatomical concepts.

Retrieval

Automatic Recall Machines: Internal Replay, Continual Learning and the Brain

1 code implementation22 Jun 2020 Xu Ji, Joao Henriques, Tinne Tuytelaars, Andrea Vedaldi

Replay in neural networks involves training on sequential data with memorized samples, which counteracts forgetting of previous behavior caused by non-stationarity.

Continual Learning

Unsupervised Model Personalization while Preserving Privacy and Scalability: An Open Problem

1 code implementation CVPR 2020 Matthias De Lange, Xu Jia, Sarah Parisot, Ales Leonardis, Gregory Slabaugh, Tinne Tuytelaars

This framework flexibly disentangles user-adaptation into model personalization on the server and local data regularization on the user device, with desirable properties regarding scalability and privacy constraints.

Continual Learning Domain Adaptation +2

Deep-Geometric 6 DoF Localization from a Single Image in Topo-metric Maps

no code implementations4 Feb 2020 Tom Roussel, Punarjay Chakravarty, Gaurav Pandey, Tinne Tuytelaars, Luc Van Eycken

We describe a Deep-Geometric Localizer that is able to estimate the full 6 Degree of Freedom (DoF) global pose of the camera from a single image in a previously mapped environment.

Pose Estimation

Ternary Feature Masks: zero-forgetting for task-incremental learning

no code implementations23 Jan 2020 Marc Masana, Tinne Tuytelaars, Joost Van de Weijer

To allow already learned features to adapt to the current task without changing the behavior of these features for previous tasks, we introduce task-specific feature normalization.

Continual Learning Incremental Learning

Information Compensation for Deep Conditional Generative Networks

no code implementations23 Jan 2020 Zehao Wang, Kaili Wang, Tinne Tuytelaars, Jose Oramas

In recent years, unsupervised/weakly-supervised conditional generative adversarial networks (GANs) have achieved many successes on the task of modeling and generating data.

Disentanglement

Online Continual Learning with Maximal Interfered Retrieval

2 code implementations NeurIPS 2019 Rahaf Aljundi, Eugene Belilovsky, Tinne Tuytelaars, Laurent Charlin, Massimo Caccia, Min Lin, Lucas Page-Caccia

Methods based on replay, either generative or from a stored memory, have been shown to be effective approaches for continual learning, matching or exceeding the state of the art in a number of standard benchmarks.

Class Incremental Learning Retrieval

How to improve CNN-based 6-DoF camera pose estimation

no code implementations23 Sep 2019 Soroush Seifi, Tinne Tuytelaars

Convolutional neural networks (CNNs) and transfer learning have recently been used for 6 degrees of freedom (6-DoF) camera pose estimation.

Data Augmentation Pose Estimation +1

A continual learning survey: Defying forgetting in classification tasks

1 code implementation18 Sep 2019 Matthias De Lange, Rahaf Aljundi, Marc Masana, Sarah Parisot, Xu Jia, Ales Leonardis, Gregory Slabaugh, Tinne Tuytelaars

Artificial neural networks thrive in solving the classification problem for a particular rigid task, acquiring knowledge through generalized learning behaviour from a distinct training phase.

Classification Continual Learning +2

Online Continual Learning with Maximally Interfered Retrieval

1 code implementation11 Aug 2019 Rahaf Aljundi, Lucas Caccia, Eugene Belilovsky, Massimo Caccia, Min Lin, Laurent Charlin, Tinne Tuytelaars

Methods based on replay, either generative or from a stored memory, have been shown to be effective approaches for continual learning, matching or exceeding the state of the art in a number of standard benchmarks.

Continual Learning Retrieval

Exploring the Challenges towards Lifelong Fact Learning

no code implementations26 Dec 2018 Mohamed Elhoseiny, Francesca Babiloni, Rahaf Aljundi, Marcus Rohrbach, Manohar Paluri, Tinne Tuytelaars

So far life-long learning (LLL) has been studied in relatively small-scale and relatively artificial setups.

Task-Free Continual Learning

1 code implementation CVPR 2019 Rahaf Aljundi, Klaas Kelchtermans, Tinne Tuytelaars

A sequence of tasks is learned, one at a time, with all data of current task available but not of previous or future tasks.

Continual Learning Face Recognition +1

Selfless Sequential Learning

1 code implementation ICLR 2019 Rahaf Aljundi, Marcus Rohrbach, Tinne Tuytelaars

In particular, we propose a novel regularizer, that encourages representation sparsity by means of neural inhibition.

Exemplar Guided Unsupervised Image-to-Image Translation with Semantic Consistency

no code implementations ICLR 2019 Liqian Ma, Xu Jia, Stamatios Georgoulis, Tinne Tuytelaars, Luc van Gool

Experimental results on various datasets show that EGSC-IT does not only translate the source image to diverse instances in the target domain, but also preserves the semantic consistency during the process.

Translation Unsupervised Image-To-Image Translation

Visual Explanation by Interpretation: Improving Visual Feedback Capabilities of Deep Neural Networks

no code implementations ICLR 2019 Jose Oramas, Kaili Wang, Tinne Tuytelaars

In this paper, we propose a novel scheme for both interpretation as well as explanation in which, given a pretrained model, we automatically identify internal features relevant for the set of classes considered by the model, without relying on additional annotations.

Super-Resolution with Deep Adaptive Image Resampling

no code implementations18 Dec 2017 Xu Jia, Hong Chang, Tinne Tuytelaars

In this work, we revisit the more traditional interpolation-based methods, that were popular before, now with the help of deep learning.

Image Super-Resolution

Memory Aware Synapses: Learning what (not) to forget

3 code implementations ECCV 2018 Rahaf Aljundi, Francesca Babiloni, Mohamed Elhoseiny, Marcus Rohrbach, Tinne Tuytelaars

We show state-of-the-art performance and, for the first time, the ability to adapt the importance of the parameters based on unlabeled data towards what the network needs (not) to forget, which may vary depending on test conditions.

Object Recognition

An Analysis of Human-centered Geolocation

2 code implementations10 Jul 2017 Kaili Wang, Yu-Hui Huang, Jose Oramas, Luc van Gool, Tinne Tuytelaars

Experiments on the Fashion 144k and a Pinterest-based dataset show that the automatic methods succeed at this task to a reasonable extent.

Pose Guided Person Image Generation

2 code implementations NeurIPS 2017 Liqian Ma, Xu Jia, Qianru Sun, Bernt Schiele, Tinne Tuytelaars, Luc van Gool

This paper proposes the novel Pose Guided Person Generation Network (PG$^2$) that allows to synthesize person images in arbitrary poses, based on an image of that person and a novel pose.

Gesture-to-Gesture Translation Pose Transfer

Speech-Based Visual Question Answering

1 code implementation1 May 2017 Ted Zhang, Dengxin Dai, Tinne Tuytelaars, Marie-Francine Moens, Luc van Gool

This paper introduces speech-based visual question answering (VQA), the task of generating an answer given an image and a spoken question.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +3

Context-based Object Viewpoint Estimation: A 2D Relational Approach

no code implementations21 Apr 2017 Jose Oramas, Luc De Raedt, Tinne Tuytelaars

To estimate the viewpoint (or pose) of an object, people have mostly looked at object intrinsic features, such as shape or appearance.

Action Recognition Object +4

Encoder Based Lifelong Learning

no code implementations ICCV 2017 Amal Rannen Triki, Rahaf Aljundi, Mathew B. Blaschko, Tinne Tuytelaars

This paper introduces a new lifelong learning solution where a single model is trained for a sequence of tasks.

Image Classification

How hard is it to cross the room? -- Training (Recurrent) Neural Networks to steer a UAV

no code implementations24 Feb 2017 Klaas Kelchtermans, Tinne Tuytelaars

To cope with more complex tasks, we propose the use of recurrent neural networks (RNN) instead and successfully train an LSTM (Long-Short Term Memory) network for controlling UAVs.

Imitation Learning

Expert Gate: Lifelong Learning with a Network of Experts

2 code implementations CVPR 2017 Rahaf Aljundi, Punarjay Chakravarty, Tinne Tuytelaars

Further, the autoencoders inherently capture the relatedness of one task to another, based on which the most relevant prior model to be used for training a new expert, with finetuning or learning without-forgetting, can be selected.

Image Classification Video Prediction

DeepProposals: Hunting Objects and Actions by Cascading Deep Convolutional Layers

1 code implementation15 Jun 2016 Amir Ghodrati, Ali Diba, Marco Pedersoli, Tinne Tuytelaars, Luc van Gool

In this paper, a new method for generating object and action proposals in images and videos is proposed.

Object

Dynamic Filter Networks

1 code implementation NeurIPS 2016 Bert De Brabandere, Xu Jia, Tinne Tuytelaars, Luc van Gool

In a traditional convolutional layer, the learned filters stay fixed after training.

 Ranked #1 on Video Prediction on KTH (Cond metric)

Depth Estimation Optical Flow Estimation +1

Online Action Detection

no code implementations21 Apr 2016 Roeland De Geest, Efstratios Gavves, Amir Ghodrati, Zhenyang Li, Cees Snoek, Tinne Tuytelaars

Third, the start of the action is unknown, so it is unclear over what time window the information should be integrated.

Online Action Detection

Modeling Visual Compatibility through Hierarchical Mid-level Elements

no code implementations31 Mar 2016 Jose Oramas, Tinne Tuytelaars

At the base-level, our method identifies patterns of CNN activations with the aim of modeling different variations/styles in which objects of the classes of interest may occur.

Object

Cross-modal Supervision for Learning Active Speaker Detection in Video

no code implementations29 Mar 2016 Punarjay Chakravarty, Tinne Tuytelaars

We further improve a generic model for active speaker detection by learning person specific models.

Action Detection Activity Detection

DeLight-Net: Decomposing Reflectance Maps into Specular Materials and Natural Illumination

no code implementations27 Mar 2016 Stamatios Georgoulis, Konstantinos Rematas, Tobias Ritschel, Mario Fritz, Luc van Gool, Tinne Tuytelaars

In this paper we are extracting surface reflectance and natural environmental illumination from a reflectance map, i. e. from a single 2D image of a sphere of one material under one illumination.

Unsupervised Domain Adaptation in the Wild: Dealing with Asymmetric Label Sets

no code implementations26 Mar 2016 Ayush Mittal, Anant Raj, Vinay P. Namboodiri, Tinne Tuytelaars

Most methods for unsupervised domain adaptation proposed in the literature to date, assume that the set of classes present in the target domain is identical to the set of classes present in the source domain.

General Classification Unsupervised Domain Adaptation

Lightweight Unsupervised Domain Adaptation by Convolutional Filter Reconstruction

no code implementations23 Mar 2016 Rahaf Aljundi, Tinne Tuytelaars

To this end, we first analyze the output of each convolutional layer from a domain adaptation perspective.

Unsupervised Domain Adaptation

Novel Views of Objects from a Single Image

no code implementations31 Jan 2016 Konstantinos Rematas, Chuong Nguyen, Tobias Ritschel, Mario Fritz, Tinne Tuytelaars

We propose a technique to use the structural information extracted from a 3D model that matches the image object in terms of viewpoint and shape.

Novel View Synthesis Object

Rank Pooling for Action Recognition

1 code implementation6 Dec 2015 Basura Fernando, Efstratios Gavves, Jose Oramas, Amir Ghodrati, Tinne Tuytelaars

We show how the parameters of a function that has been fit to the video data can serve as a robust new video representation.

Action Recognition Gesture Recognition +2

Learning to Rank Based on Subsequences

no code implementations ICCV 2015 Basura Fernando, Efstratios Gavves, Damien Muselet, Tinne Tuytelaars

We present a supervised learning to rank algorithm that effectively orders images by exploiting the structure in image sequences.

Learning-To-Rank

Continuous Pose Estimation With a Spatial Ensemble of Fisher Regressors

no code implementations ICCV 2015 Michele Fenzi, Laura Leal-Taixe, Jorn Ostermann, Tinne Tuytelaars

In this paper, we treat the problem of continuous pose estimation for object categories as a regression problem on the basis of only 2D training information.

Pose Estimation regression

Learning Where to Position Parts in 3D

no code implementations ICCV 2015 Marco Pedersoli, Tinne Tuytelaars

In this paper we propose a new method for the detection and pose estimation of 3D objects, that does not use any 3D CAD model or other 3D information.

Object object-detection +3

MidRank: Learning to rank based on subsequences

no code implementations29 Nov 2015 Basura Fernando, Efstratios Gavves, Damien Muselet, Tinne Tuytelaars

We present a supervised learning to rank algorithm that effectively orders images by exploiting the structure in image sequences.

Learning-To-Rank

Towards Automatic Image Editing: Learning to See another You

no code implementations26 Nov 2015 Amir Ghodrati, Xu Jia, Marco Pedersoli, Tinne Tuytelaars

Learning the distribution of images in order to generate new samples is a challenging task due to the high dimensionality of the data and the highly non-linear relations that are involved.

Attribute Image Generation +1

Deep Reflectance Maps

no code implementations CVPR 2016 Konstantinos Rematas, Tobias Ritschel, Mario Fritz, Efstratios Gavves, Tinne Tuytelaars

Undoing the image formation process and therefore decomposing appearance into its intrinsic properties is a challenging task due to the under-constraint nature of this inverse problem.

DeepProposal: Hunting Objects by Cascading Deep Convolutional Layers

1 code implementation ICCV 2015 Amir Ghodrati, Ali Diba, Marco Pedersoli, Tinne Tuytelaars, Luc van Gool

We generate hypotheses in a sliding-window fashion over different activation layers and show that the final convolutional layers can find the object of interest with high recall but poor localization due to the coarseness of the feature maps.

Object

Active Transfer Learning with Zero-Shot Priors: Reusing Past Datasets for Future Tasks

no code implementations ICCV 2015 Efstratios Gavves, Thomas Mensink, Tatiana Tommasi, Cees G. M. Snoek, Tinne Tuytelaars

How can we reuse existing knowledge, in the form of available datasets, when solving a new and apparently unrelated target task from a set of unlabeled data?

Active Learning General Classification +2

Subspace Alignment Based Domain Adaptation for RCNN Detector

no code implementations20 Jul 2015 Anant Raj, Vinay P. Namboodiri, Tinne Tuytelaars

In this paper, we propose subspace alignment based domain adaptation of the state of the art RCNN based object detector.

Object object-detection +2

Dataset Fingerprints: Exploring Image Collections Through Data Mining

no code implementations CVPR 2015 Konstantinos Rematas, Basura Fernando, Frank Dellaert, Tinne Tuytelaars

As the amount of visual data increases, so does the need for summarization tools that can be used to explore large image collections and to quickly get familiar with their content.

Modeling Video Evolution for Action Recognition

no code implementations CVPR 2015 Basura Fernando, Efstratios Gavves, Jose Oramas M., Amir Ghodrati, Tinne Tuytelaars

We postulate that a function capable of ordering the frames of a video temporally (based on the appearance) captures well the evolution of the appearance within the video.

Action Recognition Skeleton Based Action Recognition +1

Weakly Supervised Object Detection With Convex Clustering

no code implementations CVPR 2015 Hakan Bilen, Marco Pedersoli, Tinne Tuytelaars

However, as learning appearance and localization are two interconnected tasks, the optimization is not convex and the procedure can easily get stuck in a poor local minimum, the algorithm "misses" the object in some images.

Clustering Object +2

A Deeper Look at Dataset Bias

no code implementations6 May 2015 Tatiana Tommasi, Novi Patricia, Barbara Caputo, Tinne Tuytelaars

The presence of a bias in each image data collection has recently attracted a lot of attention in the computer vision community showing the limits in generalization of any learning method trained on a specific dataset.

Mind the Gap: Subspace based Hierarchical Domain Adaptation

no code implementations16 Jan 2015 Anant Raj, Vinay P. Namboodiri, Tinne Tuytelaars

Domain adaptation techniques aim at adapting a classifier learnt on a source domain to work on the target domain.

Domain Adaptation

Joint cross-domain classification and subspace learning for unsupervised adaptation

no code implementations17 Nov 2014 Basura Fernando, Tatiana Tommasi, Tinne Tuytelaars

Domain adaptation aims at adapting the knowledge acquired on a source domain to a new different but related target domain.

Domain Adaptation domain classification +1

Location Recognition Over Large Time Lags

no code implementations26 Sep 2014 Basura Fernando, Tatiana Tommasi, Tinne Tuytelaars

Would it be possible to automatically associate ancient pictures to modern ones and create fancy cultural heritage city maps?

Domain Adaptation

Subspace Alignment For Domain Adaptation

no code implementations18 Sep 2014 Basura Fernando, Amaury Habrard, Marc Sebban, Tinne Tuytelaars

We present two approaches to determine the only hyper-parameter in our method corresponding to the size of the subspaces.

Domain Adaptation

Object Classification with Adaptable Regions

no code implementations CVPR 2014 Hakan Bilen, Marco Pedersoli, Vinay P. Namboodiri, Tinne Tuytelaars, Luc van Gool

In classification of objects substantial work has gone into improving the low level representation of an image by considering various aspects such as different features, a number of feature pooling and coding techniques and considering different kernels.

Classification General Classification +1

Image-based Synthesis and Re-Synthesis of Viewpoints Guided by 3D Models

no code implementations CVPR 2014 Konstantinos Rematas, Tobias Ritschel, Mario Fritz, Tinne Tuytelaars

We propose a technique to use the structural information extracted from a set of 3D models of an object class to improve novel-view synthesis for images showing unknown instances of this class.

Novel View Synthesis Position +1

Using a Deformation Field Model for Localizing Faces and Facial Points under Weak Supervision

no code implementations CVPR 2014 Marco Pedersoli, Tinne Tuytelaars, Luc van Gool

Additionally, without any facial point annotation at the level of individual training images, our method can localize facial points with an accuracy similar to fully supervised approaches.

Face Detection

A Testbed for Cross-Dataset Analysis

no code implementations24 Feb 2014 Tatiana Tommasi, Tinne Tuytelaars, Barbara Caputo

Since its beginning visual recognition research has tried to capture the huge variability of the visual world in several image collections.

Seeking the Strongest Rigid Detector

no code implementations CVPR 2013 Rodrigo Benenson, Markus Mathias, Tinne Tuytelaars, Luc van Gool

The current state of the art solutions for object detection describe each class by a set of models trained on discovered sub-classes (so called "components"), with each model itself composed of collections of interrelated parts (deformable models).

feature selection object-detection +1

Cannot find the paper you are looking for? You can Submit a new open access paper.