arXiv subjects

Yosi Keller

Publications and source records attributed to Yosi Keller.

At least 19 recordsLinked to original sources

arXiv2026

LaVPR: Benchmarking Language and Vision for Place Recognition

Visual Place Recognition (VPR) often fails under extreme environmental changes and perceptual aliasing. Beyond these limitations, standard systems cannot perform 'blind' localization from verbal descriptions alone, a capability critical for applications such as emergency response. To address these challenges, we introduce LaVPR, a large-scale benchmark that extends existing VPR datasets with over 650,000 rich natural-language descriptions. Using LaVPR, we investigate two paradigms: Multi-Modal Fusion for enhanced robustness and Cross-Modal Retrieval for language-based localization. Our results show that language descriptions yield consistent gains in visually degraded conditions, with the most significant impact on smaller backbones. Notably, adding language allows compact models to rival the performance of much larger vision-only architectures. For cross-modal retrieval, we establish a baseline using Low-Rank Adaptation (LoRA) and Multi-Similarity loss, which substantially outperforms standard contrastive methods across vision-language models. Ultimately, LaVPR enables a new class of localization systems that are both resilient to real-world stochasticity and practical for resource-constrained deployment. Our dataset and code are available at https://github.com/oferidan1/LaVPR

cs.CV↗

arXiv2026

GazeShift: Unsupervised Gaze Estimation and Dataset for VR

Gaze estimation is instrumental in modern virtual reality (VR) systems. Despite significant progress in remote-camera gaze estimation, VR gaze research remains constrained by data scarcity, particularly the lack of large-scale, accurately labeled datasets captured with the off-axis camera configurations typical of modern headsets. Gaze annotation is difficult since fixation on intended targets cannot be guaranteed. To address these challenges, we introduce VRGaze, the first large-scale off-axis gaze estimation dataset for VR, comprising 2.1 million near-eye infrared images collected from 68 participants. We further propose GazeShift, an attention-guided unsupervised framework for learning gaze representations without labeled data. Unlike prior redirection-based methods that rely on multi-view or 3D geometry, GazeShift is tailored to near-eye imagery, achieving effective gaze-appearance disentanglement in a compact, real-time model. GazeShift embeddings can be optionally adapted to individual users via lightweight few-shot calibration, achieving a 1.84° mean error on VRGaze. On the remote-camera MPIIGaze dataset, the model achieves a 7.15° person-agnostic error, doing so with 10x fewer parameters and 35x fewer FLOPs than baseline methods. Deployed natively on a VR headset GPU, inference takes only 5 ms. Combined with demonstrated robustness to illumination changes, these results highlight GazeShift as a label-efficient, real-time solution for VR gaze tracking. Project code and the VRGaze dataset are released at https://github.com/gazeshift3/gazeshift

cs.CV↗

arXiv2026

Multi-Sensor Matching with HyperNetworks

Hypernetworks are models that generate or modulate the weights of another network. They provide a flexible mechanism for injecting context and task conditioning and have proven broadly useful across diverse applications without significant increases in model size. We leverage hypernetworks to improve multimodal patch matching by introducing a lightweight descriptor-learning architecture that augments a Siamese CNN with (i) hypernetwork modules that compute adaptive, per-channel scaling and shifting and (ii) conditional instance normalization that provides modality-specific adaptation (e.g., visible vs. infrared, VIS-IR) in shallow layers. This combination preserves the efficiency of descriptor-based methods during inference while increasing robustness to appearance shifts. Trained with a triplet loss and hard-negative mining, our approach achieves state-of-the-art results on VIS-NIR and other VIS-IR benchmarks and matches or surpasses prior methods on additional datasets, despite their higher inference cost. To spur progress on domain shift, we also release GAP-VIR, a cross-platform (ground/aerial) VIS-IR patch dataset with 500K pairs, enabling rigorous evaluation of cross-domain generalization and adaptation.

cs.CV↗

arXiv2025

Relative Pose Regression with Pose Auto-Encoders: Enhancing Accuracy and Data Efficiency for Retail Applications

Accurate camera localization is crucial for modern retail environments, enabling enhanced customer experiences, streamlined inventory management, and autonomous operations. While Absolute Pose Regression (APR) from a single image offers a promising solution, approaches that incorporate visual and spatial scene priors tend to achieve higher accuracy. Camera Pose Auto-Encoders (PAEs) have recently been introduced to embed such priors into APR. In this work, we extend PAEs to the task of Relative Pose Regression (RPR) and propose a novel re-localization scheme that refines APR predictions using PAE-based RPR, without requiring additional storage of images or pose data. We first introduce PAE-based RPR and establish its effectiveness by comparing it with image-based RPR models of equivalent architectures. We then demonstrate that our refinement strategy, driven by a PAE-based RPR, enhances APR localization accuracy on indoor benchmarks. Notably, our method is shown to achieve competitive performance even when trained with only 30% of the data, substantially reducing the data collection burden for retail deployment. Our code and pre-trained models are available at: https://github.com/yolish/camera-pose-auto-encoders

cs.CV↗

arXiv2025

HyperPose: Hypernetwork-Infused Camera Pose Localization and an Extended Cambridge Landmarks Dataset

In this work, we propose HyperPose, which utilizes hyper-networks in absolute camera pose regressors. The inherent appearance variations in natural scenes, attributable to environmental conditions, perspective, and lighting, induce a significant domain disparity between the training and test datasets. This disparity degrades the precision of contemporary localization networks. To mitigate this, we advocate for incorporating hypernetworks into single-scene and multiscene camera pose regression models. During inference, the hypernetwork dynamically computes adaptive weights for the localization regression heads based on the particular input image, effectively narrowing the domain gap. Using indoor and outdoor datasets, we evaluate the HyperPose methodology across multiple established absolute pose regression architectures. We also introduce and share the Extended Cambridge Landmarks (ECL), a novel localization dataset, based on the Cambridge Landmarks dataset, showing it in multiple seasons with significantly varying appearance conditions. Our empirical experiments demonstrate that HyperPose yields notable performance enhancements for single- and multi-scene architectures. We have made our source code, pre-trained models, and the ECL dataset openly available.

cs.CV↗

arXiv2025

Relative Age Estimation Using Face Images

This work introduces a novel deep-learning approach for estimating age from a single facial image by refining an initial age estimate. The refinement leverages a reference face database of individuals with similar ages and appearances. We employ a network that estimates age differences between an input image and reference images with known ages, thus refining the initial estimate. Our method explicitly models age-dependent facial variations using differential regression, yielding improved accuracy compared to conventional absolute age estimation. Additionally, we introduce an age augmentation scheme that iteratively refines initial age estimates by modeling their error distribution during training. This iterative approach further enhances the initial estimates. Our approach surpasses existing methods, achieving state-of-the-art accuracy on the MORPH II and CACD datasets. Furthermore, we examine the biases inherent in contemporary state-of-the-art age estimation techniques.

cs.CV↗

arXiv2024

Estimating Extreme 3D Image Rotation with Transformer Cross-Attention

The estimation of large and extreme image rotation plays a key role in multiple computer vision domains, where the rotated images are related by a limited or a non-overlapping field of view. Contemporary approaches apply convolutional neural networks to compute a 4D correlation volume to estimate the relative rotation between image pairs. In this work, we propose a cross-attention-based approach that utilizes CNN feature maps and a Transformer-Encoder, to compute the cross-attention between the activation maps of the image pairs, which is shown to be an improved equivalent of the 4D correlation volume, used in previous works. In the suggested approach, higher attention scores are associated with image regions that encode visual cues of rotation. Our approach is end-to-end trainable and optimizes a simple regression loss. It is experimentally shown to outperform contemporary state-of-the-art schemes when applied to commonly used image rotation datasets and benchmarks, and establishes a new state-of-the-art accuracy on these datasets. We make our code publicly available.

cs.CV↗

arXiv2023

FaceCoresetNet: Differentiable Coresets for Face Set Recognition

In set-based face recognition, we aim to compute the most discriminative descriptor from an unbounded set of images and videos showing a single person. A discriminative descriptor balances two policies when aggregating information from a given set. The first is a quality-based policy: emphasizing high-quality and down-weighting low-quality images. The second is a diversity-based policy: emphasizing unique images in the set and down-weighting multiple occurrences of similar images as found in video clips which can overwhelm the set representation. This work frames face-set representation as a differentiable coreset selection problem. Our model learns how to select a small coreset of the input set that balances quality and diversity policies using a learned metric parameterized by the face quality, optimized end-to-end. The selection process is a differentiable farthest-point sampling (FPS) realized by approximating the non-differentiable Argmax operation with differentiable sampling from the Gumbel-Softmax distribution of distances. The small coreset is later used as queries in a self and cross-attention architecture to enrich the descriptor with information from the whole set. Our model is order-invariant and linear in the input set size. We set a new SOTA to set face verification on the IJB-B and IJB-C datasets. Our code is publicly available.

cs.CV↗

arXiv2023

Hierarchical Attention-based Age Estimation and Bias Estimation

In this work we propose a novel deep-learning approach for age estimation based on face images. We first introduce a dual image augmentation-aggregation approach based on attention. This allows the network to jointly utilize multiple face image augmentations whose embeddings are aggregated by a Transformer-Encoder. The resulting aggregated embedding is shown to better encode the face image attributes. We then propose a probabilistic hierarchical regression framework that combines a discrete probabilistic estimate of age labels, with a corresponding ensemble of regressors. Each regressor is particularly adapted and trained to refine the probabilistic estimate over a range of ages. Our scheme is shown to outperform contemporary schemes and provide a new state-of-the-art age estimation accuracy, when applied to the MORPH II dataset for age estimation. Last, we introduce a bias analysis of state-of-the-art age estimation results.

cs.CV↗

arXiv2023

Attention-Based Multimodal Image Matching

We propose an attention-based approach for multimodal image patch matching using a Transformer encoder attending to the feature maps of a multiscale Siamese CNN. Our encoder is shown to efficiently aggregate multiscale image embeddings while emphasizing task-specific appearance-invariant image cues. We also introduce an attention-residual architecture, using a residual connection bypassing the encoder. This additional learning signal facilitates end-to-end training from scratch. Our approach is experimentally shown to achieve new state-of-the-art accuracy on both multimodal and single modality benchmarks, illustrating its general applicability. To the best of our knowledge, this is the first successful implementation of the Transformer encoder architecture to the multimodal image patch matching task.

cs.CV↗

arXiv2023

Coarse-to-Fine Multi-Scene Pose Regression with Transformers

Absolute camera pose regressors estimate the position and orientation of a camera given the captured image alone. Typically, a convolutional backbone with a multi-layer perceptron (MLP) head is trained using images and pose labels to embed a single reference scene at a time. Recently, this scheme was extended to learn multiple scenes by replacing the MLP head with a set of fully connected layers. In this work, we propose to learn multi-scene absolute camera pose regression with Transformers, where encoders are used to aggregate activation maps with self-attention and decoders transform latent features and scenes encoding into pose predictions. This allows our model to focus on general features that are informative for localization, while embedding multiple scenes in parallel. We extend our previous MS-Transformer approach \cite{shavit2021learning} by introducing a mixed classification-regression architecture that improves the localization accuracy. Our method is evaluated on commonly benchmark indoor and outdoor datasets and has been shown to exceed both multi-scene and state-of-the-art single-scene absolute pose regressors.

cs.CV↗

arXiv2023

Image Segmentation via Probabilistic Graph Matching

This work presents an unsupervised and semi-automatic image segmentation approach where we formulate the segmentation as a inference problem based on unary and pairwise assignment probabilities computed using low-level image cues. The inference is solved via a probabilistic graph matching scheme, which allows rigorous incorporation of low level image cues and automatic tuning of parameters. The proposed scheme is experimentally shown to compare favorably with contemporary semi-supervised and unsupervised image segmentation schemes, when applied to contemporary state-of-the-art image sets.

cs.CV↗

arXiv2023

Age-Invariant Face Embedding using the Wasserstein Distance

In this work, we study face verification in datasets where images of the same individuals exhibit significant age differences. This poses a major challenge for current face recognition and verification techniques. To address this issue, we propose a novel approach that utilizes multitask learning and a Wasserstein distance discriminator to disentangle age and identity embeddings of facial images. Our approach employs multitask learning with a Wasserstein distance discriminator that minimizes the mutual information between the age and identity embeddings by minimizing the Jensen-Shannon divergence. This improves the encoding of age and identity information in face images and enhances the performance of face verification in age-variant datasets. We evaluate the effectiveness of our approach using multiple age-variant face datasets and demonstrate its superiority over state-of-the-art methods in terms of face verification accuracy.

cs.CV↗

arXiv2023

Deep Convolutional Tables: Deep Learning without Convolutions

We propose a novel formulation of deep networks that do not use dot-product neurons and rely on a hierarchy of voting tables instead, denoted as Convolutional Tables (CT), to enable accelerated CPU-based inference. Convolutional layers are the most time-consuming bottleneck in contemporary deep learning techniques, severely limiting their use in Internet of Things and CPU-based devices. The proposed CT performs a fern operation at each image location: it encodes the location environment into a binary index and uses the index to retrieve the desired local output from a table. The results of multiple tables are combined to derive the final output. The computational complexity of a CT transformation is independent of the patch (filter) size and grows gracefully with the number of channels, outperforming comparable convolutional layers. It is shown to have a better capacity:compute ratio than dot-product neurons, and that deep CT networks exhibit a universal approximation property similar to neural networks. As the transformation involves computing discrete indices, we derive a soft relaxation and gradient-based approach for training the CT hierarchy. Deep CT networks have been experimentally shown to have accuracy comparable to that of CNNs of similar architectures. In the low compute regime, they enable an error:speed trade-off superior to alternative efficient CNN architectures.

cs.CV↗

arXiv2023

Synthesis and Reconstruction of Fingerprints using Generative Adversarial Networks

Deep learning-based models have been shown to improve the accuracy of fingerprint recognition. While these algorithms show exceptional performance, they require large-scale fingerprint datasets for training and evaluation. In this work, we propose a novel fingerprint synthesis and reconstruction framework based on the StyleGan2 architecture, to address the privacy issues related to the acquisition of such large-scale datasets. We also derive a computational approach to modify the attributes of the generated fingerprint while preserving their identity. This allows synthesizing multiple different fingerprint images per finger. In particular, we introduce the SynFing synthetic fingerprints dataset consisting of 100K image pairs, each pair corresponding to the same identity. The proposed framework was experimentally shown to outperform contemporary state-of-the-art approaches for both fingerprint synthesis and reconstruction. It significantly improved the realism of the generated fingerprints, both visually and in terms of their ability to spoof fingerprint-based verification systems. The code and fingerprints dataset are publicly available: https://github.com/rafaelbou/fingerprint_generator.

cs.CV↗

arXiv2023

Learning to Localize in Unseen Scenes with Relative Pose Regressors

Relative pose regressors (RPRs) localize a camera by estimating its relative translation and rotation to a pose-labelled reference. Unlike scene coordinate regression and absolute pose regression methods, which learn absolute scene parameters, RPRs can (theoretically) localize in unseen environments, since they only learn the residual pose between camera pairs. In practice, however, the performance of RPRs is significantly degraded in unseen scenes. In this work, we propose to aggregate paired feature maps into latent codes, instead of operating on global image descriptors, in order to improve the generalization of RPRs. We implement aggregation with concatenation, projection, and attention operations (Transformer Encoders) and learn to regress the relative pose parameters from the resulting latent codes. We further make use of a recently proposed continuous representation of rotation matrices, which alleviates the limitations of the commonly used quaternions. Compared to state-of-the-art RPRs, our model is shown to localize significantly better in unseen environments, across both indoor and outdoor benchmarks, while maintaining competitive performance in seen scenes. We validate our findings and architecture design through multiple ablations. Our code and pretrained models is publicly available.

cs.CV↗

arXiv2022

Learning to embed semantic similarity for joint image-text retrieval

We present a deep learning approach for learning the joint semantic embeddings of images and captions in a Euclidean space, such that the semantic similarity is approximated by the L2 distances in the embedding space. For that, we introduce a metric learning scheme that utilizes multitask learning to learn the embedding of identical semantic concepts using a center loss. By introducing a differentiable quantization scheme into the end-to-end trainable network, we derive a semantic embedding of semantically similar concepts in Euclidean space. We also propose a novel metric learning formulation using an adaptive margin hinge loss, that is refined during the training phase. The proposed scheme was applied to the MS-COCO, Flicke30K and Flickr8K datasets, and was shown to compare favorably with contemporary state-of-the-art approaches.

cs.CV↗

arXiv2022

Camera Pose Auto-Encoders for Improving Pose Regression

Absolute pose regressor (APR) networks are trained to estimate the pose of the camera given a captured image. They compute latent image representations from which the camera position and orientation are regressed. APRs provide a different tradeoff between localization accuracy, runtime, and memory, compared to structure-based localization schemes that provide state-of-the-art accuracy. In this work, we introduce Camera Pose Auto-Encoders (PAEs), multilayer perceptrons that are trained via a Teacher-Student approach to encode camera poses using APRs as their teachers. We show that the resulting latent pose representations can closely reproduce APR performance and demonstrate their effectiveness for related tasks. Specifically, we propose a light-weight test-time optimization in which the closest train poses are encoded and used to refine camera position estimation. This procedure achieves a new state-of-the-art position accuracy for APRs, on both the CambridgeLandmarks and 7Scenes benchmarks. We also show that train images can be reconstructed from the learned pose encoding, paving the way for integrating visual information from the train set at a low memory cost. Our code and pre-trained models are available at https://github.com/yolish/camera-pose-auto-encoders.

cs.CV↗