Search Results for author:
Found 96 papers, 48 papers with code
Leveraging Foundation models for Unsupervised Audio-Visual Segmentation
Particularly, in situations where existing supervised AVS methods struggle with overlapping foreground objects, our models still excel in accurately segmenting overlapped auditory objects.
DiffSED: Sound Event Detection with Denoising Diffusion
In this work, we reformulate the SED problem by taking a generative learning perspective.
Actor-agnostic Multi-label Action Recognition with Multi-modal Query
Existing action recognition methods are typically actor-specific due to the intrinsic topological and apparent differences among the actors.
Ranked #1 on
Zero-Shot Action Recognition
on Charades
Retrieving-to-Answer: Zero-Shot Video Question Answering with Frozen Large Language Models
Video Question Answering (VideoQA) has been significantly advanced from the scaling of recent Large Language Models (LLMs).
Ranked #2 on
Temporal/Casual QA
on NExT-QA
(using extra training data)
HeadSculpt: Crafting 3D Head Avatars with Text
Specifically, we first equip the diffusion model with 3D awareness by leveraging landmark-based control and a learned textual embedding representing the back view appearance of heads, enabling 3D-consistent head avatar generations.
DiffTAD: Temporal Action Detection with Proposal Denoising Diffusion
Concretely, we establish the denoising process in the Transformer decoder (e. g., DETR) by introducing a temporal location query design with faster convergence in training.
Generative Semantic Segmentation
To that end, the segmentation mask is expressed with a special type of image (dubbed as maskige).
PersonalTailor: Personalizing 2D Pattern Design from 3D Garment Point Clouds
Garment pattern design aims to convert a 3D garment to the corresponding 2D panels and their sewing structure.
FAME-ViL: Multi-Tasking Vision-Language Model for Heterogeneous Fashion Tasks
In the fashion domain, there exists a variety of vision-and-language (V+L) tasks, including cross-modal retrieval, text-guided image retrieval, multi-modal classification, and image captioning.
Unsupervised Hashing with Similarity Distribution Calibration
However, these methods often overlook the fact that the similarity between data points in the continuous feature space may not be preserved in the discrete hash code space, due to the limited similarity range of hash codes.
Preconditioned Score-based Generative Models
Compared with the latest generative models (\eg, CLD-SGM, DDIM, and Analytic-DDIM), PDS can achieve the best sampling quality on CIFAR-10 at a FID score of 1. 99.
Homeomorphism Alignment for Unsupervised Domain Adaptation
With distribution alignment, it is challenging to acquire a common space which maintains fully the discriminative structure of both domains.
Controllable Person Image Synthesis with Pose-Constrained Latent Diffusion
Controllable person image synthesis aims at rendering a source image based on user-specified changes in body pose or appearance.
DeepChange: A Long-Term Person Re-Identification Benchmark with Clothes Change
Long-term re-id with clothes change is a challenging problem in surveillance AI.
Multi-Modal Few-Shot Temporal Action Detection
In this work, we introduce a new multi-modality few-shot (MMFS) TAD problem, which can be considered as a marriage of FS-TAD and ZS-TAD by leveraging few-shot support videos and new class names jointly.
Post-Processing Temporal Action Detection
To address this problem, in this work we introduce a novel model-agnostic post-processing method without model redesign and retraining.
Self-supervised Video Representation Learning with Motion-Aware Masked Autoencoders
As a result, our model can extract effectively both static appearance and dynamic motion spontaneously, leading to superior spatiotemporal representation learning capability.
Semi-Supervised and Unsupervised Deep Visual Learning: A Survey
Semi-supervised learning and unsupervised learning offer promising paradigms to learn from an abundance of unlabeled visual data.
DeepInteraction: 3D Object Detection via Modality Interaction
Existing top-performance 3D object detectors typically rely on the multi-modal fusion strategy.
KUNet: Imaging Knowledge-Inspired Single HDR Image Reconstruction
Recently, with the rise of high dynamic range (HDR) display devices, there is a great demand to transfer traditional low dynamic range (LDR) images into HDR versions.
Vision Transformers: From Semantic Segmentation to Dense Prediction
In this work, for the first time we explore the global context learning potentials of ViTs for dense visual prediction (e. g., semantic segmentation).
Zero-Shot Temporal Action Detection via Vision-Language Prompting
Such a novel design effectively eliminates the dependence between localization and classification by breaking the route for error propagation in-between.
Ranked #1 on
Zero-Shot Action Detection
on THUMOS' 14
FashionViL: Fashion-Focused Vision-and-Language Representation Learning
We thus propose a Multi-View Contrastive Learning task for pulling closer the visual representation of one image to the compositional multimodal representation of another image+text.
Semi-Supervised Temporal Action Detection with Proposal-Free Masking
Such a novel design effectively eliminates the dependence between localization and classification by cutting off the route for error propagation in-between.
Ranked #1 on
Semi-Supervised Action Detection
on THUMOS' 14
Proposal-Free Temporal Action Detection via Global Segmentation Mask Learning
Existing temporal action detection (TAD) methods rely on generating an overwhelmingly large number of proposals per video.
Ranked #10 on
Temporal Action Localization
on ActivityNet-1.3
Softmax-free Linear Transformers
With linear complexity, much longer token sequences are permitted by SOFT, resulting in superior trade-off between accuracy and complexity.
Accelerating Score-based Generative Models with Preconditioned Diffusion Sampling
However, a fundamental limitation is that their inference is very slow due to a need for many (e. g., 2000) iterations of sequential computations.
PolarFormer: Multi-camera 3D Object Detection with Polar Transformer
3D object detection in autonomous driving aims to reason "what" and "where" the objects of interest present in a 3D world.
Ranked #2 on
Robust Camera Only 3D Object Detection
on nuScenes-C
ST-Adapter: Parameter-Efficient Image-to-Video Transfer Learning
This has led to a new research direction in parameter-efficient transfer learning.
Ranked #18 on
Action Recognition
on Something-Something V2
(using extra training data)
Multimodal Learning with Transformers: A Survey
Transformer is a promising neural network learner, and has achieved great success in various machine learning tasks.
Accelerating Score-based Generative Models for High-Resolution Image Synthesis
To ensure stability of convergence in sampling and generation quality, however, this sequential sampling process has to take a small step size and many sampling iterations (e. g., 2000).
Learning Ego 3D Representation as Ray Tracing
A self-driving perception model aims to extract 3D semantic representations from multiple cameras collectively into the bird's-eye-view (BEV) coordinate frame of the ego car in order to ground downstream planner.
Knowledge Distillation Meets Open-Set Semi-Supervised Learning
The key idea is that we leverage the teacher's classifier as a semantic critic for evaluating the representations of both teacher and student and distilling the semantic knowledge with high-order structured information over all feature dimensions.
EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers
In this work, pushing further along this under-studied direction we introduce EdgeViTs, a new family of light-weight ViTs that, for the first time, enable attention-based vision models to compete with the best light-weight CNNs in the tradeoff between accuracy and on-device efficiency.
Open-Set Semi-Supervised Learning for 3D Point Cloud Understanding
We also demonstrate the feasibility of a more efficient training strategy.
SOS! Self-supervised Learning Over Sets Of Handled Objects In Egocentric Action Recognition
To overcome both limitations, we introduce Self-Supervised Learning Over Sets (SOS), an approach to pre-train a generic Objects In Contact (OIC) representation model from video object regions detected by an off-the-shelf hand-object contact detector.
Unsupervised Long-Term Person Re-Identification with Clothes Change
We investigate unsupervised person re-identification (Re-ID) with clothes change, a new challenging problem with more practical usability and scalability to real-world deployment.
Clustering
Unsupervised Long Term Person Re-Identification
+1
Source-Free Object Detection by Learning To Overlook Domain Style
This approach suffers from both unsatisfactory accuracy of pseudo labels due to the presence of domain shift and limited use of target domain training data.
Low-Fidelity Video Encoder Optimization for Temporal Action Localization
This results in a task discrepancy problem for the video encoder – trained for action classification, but used for TAL.
Ranked #6 on
Temporal Action Localization
on HACS
SOFT: Softmax-free Transformer with Linear Complexity
Crucially, with a linear complexity, much longer token sequences are permitted in SOFT, resulting in superior trade-off between accuracy and complexity.
Few-Shot Temporal Action Localization with Query Adaptive Transformer
Further, a novel FS-TAL model is proposed which maximizes the knowledge transfer from training classes whilst enabling the model to be dynamically adapted to both the new class and each video of that class simultaneously.
Ranked #1 on
Few Shot Temporal Action Localization
on ActivityNet
Action Segmentation
Few Shot Temporal Action Localization
+4
Temporal Action Localization with Global Segmentation Mask Transformers
In this paper, to address the above two challenges, a novel {\em Global Segmentation Mask Transformer} (GSMT) is proposed.
Single Person Pose Estimation: A Survey
Human pose estimation in unconstrained images and videos is a fundamental computer vision task.
Low-resolution Human Pose Estimation
Human pose estimation has achieved significant progress on images with high imaging resolution.
Simpler is Better: Few-shot Semantic Segmentation with Classifier Weight Transformer
A few-shot semantic segmentation model is typically composed of a CNN encoder, a CNN decoder and a simple classifier (separating foreground and background pixels).
Global Aggregation then Local Distribution for Scene Parsing
Modelling long-range contextual relationships is critical for pixel-wise prediction tasks such as semantic segmentation.
DeepChange: A Large Long-Term Person Re-Identification Benchmark with Clothes Change
Currently, one of the most significant limitations in this field is the lack of a large realistic benchmark.
Low-Fidelity End-to-End Video Encoder Pre-training for Temporal Action Localization
This results in a task discrepancy problem for the video encoder -- trained for action classification, but used for TAL.
Few-Shot Website Fingerprinting Attack
To address this, we introduce a model-agnostic, efficient, and Harmonious Data Augmentation (HDA) method that can improve deep WF attacking methods significantly.
Data Augmentation
Cryptography and Security
68M25
K.4.1
Few-shot Action Recognition with Prototype-centered Attentive Learning
Extensive experiments on four standard few-shot action benchmarks show that our method clearly outperforms previous state-of-the-art methods, with the improvement particularly significant (10+\%) on the most challenging fine-grained action recognition benchmark.
Unsupervised Noisy Tracklet Person Re-identification
Extensive comparative experiments demonstrate that the proposed STL model surpasses significantly the state-of-the-art unsupervised learning and one-shot learning re-id methods on three large tracklet person re-id benchmarks.
Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers
In this paper, we aim to provide an alternative perspective by treating semantic segmentation as a sequence-to-sequence prediction task.
Ranked #2 on
Semantic Segmentation
on FoodSeg103
(using extra training data)
Towards Uncovering the Intrinsic Data Structures for Unsupervised Domain Adaptation using Structurally Regularized Deep Clustering
To address this issue, we are motivated by a UDA assumption of structural similarity across domains, and propose to directly uncover the intrinsic target discrimination via constrained clustering, where we constrain the clustering solutions using structural source regularization that hinges on the very same assumption.
Boundary-sensitive Pre-training for Temporal Localization in Videos
However, most existing models developed for these tasks are pre-trained on general video action classification tasks.
Ranked #17 on
Temporal Action Localization
on ActivityNet-1.3
Egocentric Action Recognition by Video Attention and Temporal Context
In this challenge, action recognition is posed as the problem of simultaneously predicting a single `verb' and `noun' class label given an input trimmed video clip.
Knowing What, Where and When to Look: Efficient Video Action Modeling with Attention
Departing from existing alternatives, our W3 module models all three facets of video attention jointly.
Ranked #1 on
Action Recognition
on EgoGesture
Joint COCO and Mapillary Workshop at ICCV 2019 Keypoint Detection Challenge Track Technical Report: Distribution-Aware Coordinate Representation for Human Pose Estimation
Taking them together, we formulate a novel Distribution-Aware coordinate Representation for Keypoint (DARK) method.
Incremental Few-Shot Object Detection
To this end we propose OpeN-ended Centre nEt (ONCE), a detector designed for incrementally learning to detect novel class objects with few examples.
Intra-Camera Supervised Person Re-Identification
Existing person re-identification (re-id) methods mostly exploit a large set of cross-camera identity labelled training data.
Characteristic Regularisation for Super-Resolving Face Images
Extensive evaluations demonstrate the performance superiority of our method over state-of-the-art SR and UDA models on both genuine and artificial LR facial imagery data.
Distribution-Aware Coordinate Representation for Human Pose Estimation
Interestingly, we found that the process of decoding the predicted heatmaps into the final joint coordinates in the original image space is surprisingly significant for human pose estimation performance, which nevertheless was not recognised before.
Ranked #2 on
Multi-Person Pose Estimation
on COCO
(using extra training data)
Neural Operator Search
We address this limitation by additionally exploiting feature self-calibration operations, resulting in a heterogeneous search space.
Intra-Camera Supervised Person Re-Identification: A New Benchmark
Existing person re-identification (re-id) methods rely mostly on a large set of inter-camera identity labelled training data, requiring a tedious data collection and annotation process therefore leading to poor scalability in practical re-id applications.
Universal Person Re-Identification
Most state-of-the-art person re-identification (re-id) methods depend on supervised model learning with a large set of cross-view identity labelled training data.
Unsupervised Deep Learning by Neighbourhood Discovery
Deep convolutional neural networks (CNNs) have demonstrated remarkable success in computer vision by supervisedly learning strong visual feature representations.
Unsupervised Tracklet Person Re-Identification
We formulate an Unsupervised Tracklet Association Learning (UTAL) framework.
Ranked #1 on
Person Re-Identification
on DukeTracklet
Low-Resolution Face Recognition
Whilst recent face-recognition (FR) techniques have made significant progress on recognising constrained high-resolution web images, the same cannot be said on natively unconstrained low-resolution images at large scales.
Single-Label Multi-Class Image Classification by Deep Logistic Regression
The objective learning formulation is essential for the success of convolutional neural networks.
Self-Referenced Deep Learning
Whilst being able to create stronger target networks compared to the vanilla non-teacher based learning strategy, this scheme needs to train additionally a large teacher model with expensive computational cost.
Fast Human Pose Estimation
In this work, we investigate the under-studied but practically critical pose model efficiency problem.
Ranked #9 on
Pose Estimation
on Leeds Sports Poses
Vehicle Re-Identification in Context
Existing vehicle re-identification (re-id) evaluation benchmarks consider strongly artificial test scenarios by assuming the availability of high quality images and fine-grained appearance at an almost constant image scale, reminiscent to images required for Automatic Number Plate Recognition, e. g. VeRi-776.
Unsupervised Person Re-identification by Deep Learning Tracklet Association
Mostexistingpersonre-identification(re-id)methods relyon supervised model learning on per-camera-pair manually labelled pairwise training data.
Ranked #2 on
Person Re-Identification
on DukeTracklet
Semi-Supervised Deep Learning with Memory
We consider the semi-supervised multi-class classification problem of learning from sparse labelled and abundant unlabelled training data.
Deep Association Learning for Unsupervised Video Person Re-identification
In this work, to address the video person re-id task, we formulate a novel Deep Association Learning (DAL) scheme, the first end-to-end deep learning method using none of the identity labels in model initialisation and training.
Ranked #6 on
Person Re-Identification
on PRID2011
Unsupervised Person Re-Identification
Video-Based Person Re-Identification
Person Search by Multi-Scale Matching
In contrast to previous studies, we show that sufficiently reliable person instance cropping is achievable by slightly improved state-of-the-art deep learning object detectors (e. g. Faster-RCNN), and the under-studied multi-scale matching problem in person search is a more severe barrier.
Open Logo Detection Challenge
In this work, we introduce a more realistic and challenging logo detection setting, called Open Logo Detection.
Person Re-Identification in Identity Regression Space
Most existing person re-identification (re-id) methods are unsuitable for real-world deployment due to two reasons: Unscalability to large population size, and Inadaptability over time.
Knowledge Distillation by On-the-Fly Native Ensemble
Knowledge distillation is effective to train small and generalisable network models for meeting the low-memory and fast running requirements.
Imbalanced Deep Learning by Minority Class Incremental Rectification
In particular, existing deep learning methods consider mostly either class balanced data or moderately imbalanced data in model training, and ignore the challenge of learning from significantly imbalanced training data.
Surveillance Face Recognition Challenge
To facilitate more studies on developing FR models that are effective and robust for low-resolution surveillance facial images, we introduce a new Surveillance Face Recognition Challenge, which we call the QMUL-SurvFace benchmark.
Scalable Deep Learning Logo Detection
Existing logo detection methods usually consider a small number of logo classes and limited images per class with a strong assumption of requiring tedious object bounding box annotations, therefore not scalable to real-world dynamic applications.
Transferable Joint Attribute-Identity Deep Learning for Unsupervised Person Re-Identification
Most existing person re-identification (re-id) methods require supervised model learning from a separate large set of pairwise labelled training data for every single camera pair.
Ranked #22 on
Unsupervised Domain Adaptation
on Market to Duke
Unsupervised Domain Adaptation
Unsupervised Person Re-Identification
Harmonious Attention Network for Person Re-Identification
Existing person re-identification (re-id) methods either assume the availability of well-aligned person bounding box images as model input or rely on constrained attention selection mechanisms to calibrate misaligned images.
Ranked #13 on
Person Re-Identification
on CUHK03 detected
Class Rectification Hard Mining for Imbalanced Deep Learning
Recognising detailed facial or clothing attributes in images of people is a challenging task for computer vision, especially when the training data are both in very large scale and extremely imbalanced among different attribute classes.
Attribute Recognition by Joint Recurrent Learning of Context and Correlation
Recognising semantic pedestrian attributes in surveillance images is a challenging task for computer vision, particularly when the imaging quality is poor with complex background clutter and uncontrolled viewing conditions, and the number of labelled training data is small.
Multi-Label Image Classification
Pedestrian Attribute Recognition
Deep Reinforcement Learning Attention Selection for Person Re-Identification
Existing person re-identification (re-id) methods assume the provision of accurately cropped person bounding boxes with minimum background noise, mostly by manually cropping.
Discovering Visual Concept Structure with Sparse and Incomplete Tags
As a result, our model is able to discover more accurate semantic correlation between textual tags and visual features, and finally providing favourable visual semantics interpretation even with highly sparse and incomplete tags.
Person Re-Identification by Deep Joint Learning of Multi-Loss Classification
Existing person re-identification (re-id) methods rely mostly on either localised or global feature representation alone.
Ranked #99 on
Person Re-Identification
on Market-1501
Person Re-Identification by Camera Correlation Aware Feature Augmentation
The challenge of person re-identification (re-id) is to match individual images of the same person captured by different non-overlapping camera views against significant and unknown cross-view feature distortion.
Deep Learning Logo Detection with Data Expansion by Synthesising Context
Logo detection in unconstrained images is challenging, particularly when only very sparse labelled training images are accessible due to high labelling costs.
Human-In-The-Loop Person Re-Identification
Current person re-identification (re-id) methods assume that (1) pre-labelled training data are available for every camera pair, (2) the gallery size for re-identification is moderate.
Person Re-Identification by Unsupervised Video Matching
Crucially, this model does not require pairwise labelled training data (i. e. unsupervised) therefore readily scalable to large scale camera networks of arbitrary camera pairs without the need for exhaustive data annotation for every camera pair.
Multi-Task Curriculum Transfer Deep Learning of Clothing Attributes
Recognising detailed clothing characteristics (fine-grained attributes) in unconstrained images of people in-the-wild is a challenging task for computer vision, especially when there is only limited training data from the wild whilst most data available for model learning are captured in well-controlled environments using fashion models (well lit, no background clutter, frontal view, high-resolution).
Person Re-Identification by Discriminative Selection in Video Ranking
Current person re-identification (ReID) methods typically rely on single-frame imagery features, whilst ignoring space-time information from image sequences often available in the practical surveillance scenarios.
Learning from Multiple Sources for Video Summarisation
Many visual surveillance tasks, e. g. video summarisation, is conventionally accomplished through analysing imagerybased features.
Constructing Robust Affinity Graphs for Spectral Clustering
Spectral clustering requires robust and meaningful affinity graphs as input in order to form clusters with desired structures that can well support human intuition.
