Gait recognition is a computer vision task that identifies individuals based on their walking patterns. Gait recognition performance is commonly evaluated by ranking a gallery of candidates and measuring the accuracy at the top Rank-$K$. Existing models are typically single-staged, i.e. searching for the probe's nearest neighbors in a gallery using a single global feature representation. Although these models typically excel at retrieving the correct identity within the top-$K$ predictions, they struggle when hard negatives appear in the top short-list, leading to relatively low performance at the highest ranks (e.g., Rank-1). In this paper, we introduce CarGait, a Cross-Attention Re-ranking method for gait recognition, that involves re-ordering the top-$K$ list leveraging the fine-grained correlations between pairs of gait sequences through cross-attention between gait strips. This re-ranking scheme can be adapted to existing single-stage models to enhance their final results. We demonstrate the capabilities of CarGait by extensive experiments on three common gait datasets, Gait3D, GREW, and OU-MVLP, and seven different gait models, showing consistent improvements in Rank-1,5 accuracy, superior results over existing re-ranking methods, and strong baselines.

Diffusion inversion is the problem of taking an image and a text prompt that describes it and finding a noise latent that would generate the exact same image. Most current deterministic inversion techniques operate by approximately solving an implicit equation and may converge slowly or yield poor reconstructed images. We formulate the problem by finding the roots of an implicit equation and devlop a method to solve it efficiently. Our solution is based on Newton-Raphson (NR), a well-known technique in numerical analysis. We show that a vanilla application of NR is computationally infeasible while naively transforming it to a computationally tractable alternative tends to converge to out-of-distribution solutions, resulting in poor reconstruction and editing. We therefore derive an efficient guided formulation that fastly converges and provides high-quality reconstructions and editing. We showcase our method on real image editing with three popular open-sourced diffusion models: Stable Diffusion, SDXL-Turbo, and Flux with different deterministic schedulers. Our solution, **Guided Newton-Raphson Inversion**, inverts an image within 0.4 sec (on an A100 GPU) for few-step models (SDXL-Turbo and Flux.1), opening the door for interactive image editing. We further show improved results in image interpolation and generation of rare objects.

The task of Visual Place Recognition (VPR) is to predict the location of a query image from a database of geo-tagged images. Recent studies in VPR have highlighted the significant advantage of employing pre-trained foundation models like DINOv2 for the VPR task. However, these models are often deemed inadequate for VPR without further fine-tuning on VPR-specific data. In this paper, we present an effective approach to harness the potential of a foundation model for VPR. We show that features extracted from self-attention layers can act as a powerful re-ranker for VPR, even in a zero-shot setting. Our method not only outperforms previous zero-shot approaches but also introduces results competitive with several supervised methods. We then show that a single-stage approach utilizing internal ViT layers for pooling can produce global features that achieve state-of-the-art performance, with impressive feature compactness down to 128D. Moreover, integrating our local foundation features for re-ranking further widens this performance gap. Our method also demonstrates exceptional robustness and generalization, setting new state-of-the-art performance, while handling challenging conditions such as occlusion, day-night transitions, and seasonal variations.

We address the challenge of Small Object Image Retrieval (SoIR), where the goal is to retrieve images containing a specific small object, in a cluttered scene. The key challenge in this setting is constructing a single image descriptor, for scalable and efficient search, that effectively represents all objects in the image. In this paper, we first analyze the limitations of existing methods on this challenging task and then introduce new benchmarks to support SoIR evaluation. Next, we introduce \ours (\oursMI), a novel retrieval framework which incorporates a dedicated multi-object pre-training phase. This is followed by a refinement process that leverages attention-based feature extraction with object masks, integrating them into a single unified image descriptor. Our \oursMI approach significantly outperforms existing retrieval methods and strong baselines, achieving notable improvements in both zero-shot and lightweight multi-object fine-tuning. We hope this work will pave the way and inspire further research to enhance retrieval performance for this highly practical task.

In Omnimatte, one aims to decompose a given video into semantically meaningful layers, including the background and individual objects along with their associated effects, such as shadows and reflections. Existing methods often require extensive training or costly self-supervised optimization. In this paper, we present OmnimatteZero, a training-free approach that leverages off-the-shelf pre-trained video diffusion models for omnimatte. It can remove objects from videos, extract individual object layers along with their effects, and composite those objects onto new videos. These are accomplished by adapting zero-shot image inpainting techniques for video object removal, a task they fail to handle effectively out-of-the-box. To overcome this, we introduce temporal and spatial attention guidance modules that steer the diffusion process for accurate object removal and temporally consistent background reconstruction. We further show that self-attention maps capture information about the object and its footprints and use them to inpaint the object’s effects, leaving a clean background. Additionally, through simple latent arithmetic, object layers can be isolated and recombined seamlessly with new video layers to produce new videos. Evaluations show that OmnimatteZero not only achieves superior performance in terms of background reconstruction but also sets a new record for the fastest Omnimatte approach, achieving real-time performance with minimal frame runtime. Project Page.

The task of Composed Image Retrieval (CoIR) involves queries that combine image and text modalities, allowing users to express their intent more effectively. However, current CoIR datasets are orders of magnitude smaller compared to other vision and language (V&L) datasets. Additionally, some of these datasets have noticeable issues, such as queries containing redundant modalities. To address these shortcomings, we introduce the Large Scale Composed Image Retrieval (LaSCo) dataset, a new CoIR dataset which is ten times larger than existing ones. Pre-training on our LaSCo, shows a noteworthy improvement in performance, even in zero-shot. Furthermore, we propose a new approach for analyzing CoIR datasets and methods, which detects modality redundancy or necessity, in queries. We also introduce a new CoIR baseline, the Cross-Attention driven Shift Encoder (CASE). This baseline allows for early fusion of modalities using a cross-attention module and employs an additional auxiliary task during training. Our experiments demonstrate that this new baseline outperforms the current state-of-the-art methods on established benchmarks like FashionIQ and CIRR.

Text-to-image diffusion models can synthesize high quality images, but they have various limitations. Here we highlight a common failure mode of these models, namely, generating uncommon concepts and structured concepts like hand palms. We show that their limitation is partly due to the long-tail nature of their training data: web-crawled data sets are strongly unbalanced, causing models to under-represent concepts from the tail of the distribution. We characterize the effect of unbalanced training data on text-to-image models and offer a remedy. We show that rare concepts can be correctly generated by carefully selecting suitable generation seeds in the noise space, using a small reference set of images, a technique that we call SeedSelect. SeedSelect does not require retraining or finetuning the diffusion model. We assess the faithfulness, quality and diversity of SeedSelect in creating rare objects and generating complex formations like hand images, and find it consistently achieves superior performance. We further show the advantage of SeedSelect in semantic data augmentation. Generating semantically appropriate images can successfully improve performance in few-shot recognition benchmarks, for classes from the head and from the tail of the training data of diffusion models.

Personalized retrieval and segmentation aim to locate specific instances within a dataset based on an input image and a short description of the reference instance. While supervised methods are effective, they require extensive labeled data for training. Recently, self-supervised foundation models have been introduced to these tasks showing comparable results to supervised methods. However, a significant flaw in these models is evident: they struggle to locate a desired instance when other instances within the same class are presented. In this paper, we explore text-to-image diffusion models for these tasks. Specifically, we propose a novel approach called PDM for Personalized Diffusion Features Matching, that leverages intermediate features of pre-trained text-to-image models for personalization tasks without any additional training. PDM demonstrates superior performance on popular retrieval and segmentation benchmarks, outperforming even supervised methods. We also highlight notable shortcomings in current instance and segmentation datasets and propose new benchmarks for these tasks.

Gait Recognition is a computer vision task aiming to identify people by their walking patterns. Although existing methods often show high performance on specific datasets, they lack the ability to generalize to unseen scenarios. Unsupervised Domain Adaptation (UDA) tries to adapt a model, pre-trained in a supervised manner on a source domain, to an unlabelled target domain. There are only a few works on UDA for gait recognition proposing solutions to limited scenarios. In this paper, we reveal a fundamental phenomenon in adaptation of gait recognition models, caused by the bias in the target domain to viewing angle or walking direction. We then suggest a remedy to reduce this bias with a novel triplet selection strategy combined with curriculum learning. To this end, we present Gait Orientation-based method for Unsupervised Domain Adaptation (GOUDA). We provide extensive experiments on four widely-used gait datasets, CASIA-B, OU-MVLP, GREW, and Gait3D, and on three backbones, GaitSet, Gait-Part, and GaitGL, justifying the view bias and showing the superiority of our proposed method over prior UDA works.

Chats emerge as an effective user-friendly approach for information retrieval, and are successfully employed in many domains, such as customer service, healthcare, and finance. However, existing image retrieval approaches typically address the case of a single query-to-image round, and the use of chats for image retrieval has been mostly overlooked. In this work, we introduce ChatIR: a chat-based image retrieval system that engages in a conversation with the user to elicit information, in addition to an initial query, in order to clarify the user's search intent. Motivated by the capabilities of today's foundation models, we leverage Large Language Models to generate follow-up questions to an initial image description. These questions form a dialog with the user in order to retrieve the desired image from a large corpus. In this study, we explore the capabilities of such a system tested on a large dataset and reveal that engaging in a dialog yields significant gains in image retrieval. We start by building an evaluation pipeline from an existing manually generated dataset and explore different modules and training strategies for ChatIR. Our comparison includes strong baselines derived from related applications trained with Reinforcement Learning. Our system is capable of retrieving the target image from a pool of 50K images with over 78% success rate after 5 dialogue rounds, compared to 75% when questions are asked by humans, and 64% for a single shot text-to-image retrieval. Extensive evaluations reveal the strong capabilities and examine the limitations of CharIR under different settings. Project repository is available at https://github.com/levymsn/ChatIR.

Text-to-image diffusion models show great potential in synthesizing a large variety of concepts in new compositions and scenarios. However, the latent space of initial seeds is still not well understood and its structure was shown to impact the generation of various concepts. Specifically, simple operations like interpolation and finding the centroid of a set of seeds perform poorly when using standard Euclidean or spherical metrics in the latent space. This paper makes the observation that, in current training procedures, diffusion models observed inputs with a narrow range of norm values. This has strong implications for methods that rely on seed manipulation for image generation, with applications to few-shot and long-tail learning tasks. To address this issue, we propose a novel method for interpolating between two seeds and demonstrate that it defines a new non-Euclidean metric that takes into account a norm-based prior on seeds. We describe a simple yet efficient algorithm for approximating this interpolation procedure and use it to further define centroids in the latent seed space. We show that our new interpolation and centroid techniques significantly enhance the generation of rare concept images. This further leads to state-of-the-art performance on few-shot and long-tail benchmarks, improving prior approaches in terms of generation speed, image quality, and semantic content.

One of the key factors of enabling machine learning models to comprehend and solve real-world tasks is to leverage multimodal data. Unfortunately, annotation of multimodal data is challenging and expensive. Recently, self-supervised multimodal methods that combine vision and language were proposed to learn multimodal representations without annotation. However, these methods often choose to ignore the presence of high levels of noise and thus yield sub-optimal results. In this work, we show that the problem of noise estimation for multimodal data can be reduced to a multimodal density estimation task. Using multimodal density estimation, we propose a noise estimation building block for multimodal representation learning that is based strictly on the inherent correlation between different modalities. We demonstrate how our noise estimation can be broadly integrated and achieves comparable results to state-of-the-art performance on five different benchmark datasets for two challenging multimodal tasks: Video Question Answering and Text-To-Video Retrieval. Furthermore, we provide a theoretical probabilistic error bound substantiating our empirical results and analyze failure cases. Code: https://github.com/elad-amrani/ssml.

Video scene detection is the task of dividing videos into temporal semantic chapters. This is an important preliminary step before attempting to analyze heterogeneous video content. Recently, Optimal Sequential Grouping (OSG) was proposed as a powerful unsupervised solution to solve a formulation of the video scene detection problem. In this work, we extend the capabilities of OSG to the learning regime. By giving the capability to both learn from examples and leverage a robust optimization formulation, we can boost performance and enhance the versatility of the technology. We present a comprehensive analysis of incorporating OSG into deep learning neural networks under various configurations. These configurations include learning an embedding in a straight-forward manner, a tailored loss designed to guide the solution of OSG, and an integrated model where the learning is performed through the OSG pipeline. With thorough evaluation and analysis, we assess the benefits and behavior of the various configurations, and show that our learnable OSG approach exhibits desirable behavior and enhanced performance compared to the state of the art.

The flourishing of machine learning use for cognitive tasks has driven an increased demand for large annotated training datasets. In the medical imaging domain, such datasets are scarce, and the process of labeling them is costly, error prone and requires high expertise. Unsupervised learning is therefore an attractive approach for analyzing unlabeled medical images. In this paper we describe an unsupervised analysis method, consisting of feature learning by Stacked Auto-Encoders, K-means clustering for building a data model, and encoding of new images using the model. We utilize this method for image-level and patch-level analysis of breast mammograms. At the image-level, we demonstrate that our cluster-based image encoding is able to identify outlier images such as images with implants or non-standard acquisition views. At the patch-level, we show that image signatures using patch clustering can be used for unsupervised semantic segmentation of breast tissues, as well as for separating mammograms with high and low breast density. We evaluate our suggested methods on large datasets and discuss potential applications for data curation, machine-guided annotation and automatic interpretation of medical images.

Depth from Defocus (DFD) suggests a simple optical set-up to recover the shape of a scene through imaging with shallow depth of field. Although numerous methods have been proposed for DFD, less attention has been paid to the particular problem of alignment between the captured images. The inherent shift-variant defocus often prevents standard registration techniques from achieving the accuracy needed for successful shape reconstruction. In this paper, we address the DFD and registration problem in a unified framework, exploiting their mutual relation to reach a better solution for both cues. We draw a formal connection between registration and defocus blur, find its limitations and reveal the weakness of the standard isolated approaches of registration and depth estimation. The solution is approached by energy minimization. The efficiency of the associated numerical scheme is justified by showing its equivalence to the celebrated Newton-Raphson method and proof of convergence of the emerged linear system. The computationally intensive approach of DFD, newly combined with simultaneous registration, is handled by GPU computing. Experimental results demonstrate the high sensitivity of the recovered shapes to slight errors in registration and validate the superior performance of the suggested approach over two, separately applying registration and DFD alternatives.

This paper addresses the problem of correspondence establishment in binocular stereo vision. We suggest a novel spatially continuous approach for stereo matching based on the variational framework. The proposed method suggests a unique regularization term based on Mumford-Shah functional for discontinuity preserving, combined with a new energy functional for occlusion handling. The evaluation process is based on concurrent minimization of two coupled energy functionals, one for domain segmentation (occluded versus visible) and the other for disparity evaluation. In addition to a dense disparity map, our method also provides an estimation for the half-occlusion domain and a discontinuity function allocating the disparity/depth boundaries. Two new constraints are introduced improving the revealed discontinuity map. The experimental tests include a wide range of real data sets from the Middlebury stereo database. The results demonstrate the capability of our method in calculating an accurate disparity function with sharp discontinuities and occlusion map recovery. Significant improvements are shown compared to a recently published variational stereo approach. A comparison on the Middlebury stereo benchmark with subpixel accuracies shows that our method is currently among the top-ranked stereo matching algorithms.

This paper addresses the problem of correspondence establishment in binocular stereo vision. We suggest a novel variational approach that considers both the discontinuities and occlusions. It deals with color images as well as gray levels. The proposed method divides the image domain into the visible and occluded regions where each region is handled differently. The depth discontinuities in the visible domain are preserved by use of the total variation term in conjunction with the Mumford-Shah framework. In addition to the dense disparity and the occlusion maps, our method also provides a discontinuity function revealing the location of the boundaries in the disparity map. We evaluate our method on data sets from Middlebury site showing superior performance in comparison to the state of the art variational technique.

The problem of dense optical flow computation is addressed from a variational viewpoint. A new geometric framework is introduced. It unifies previous art and yields new efficient methods. Along with the framework a new alignment criterion suggests itself. It is shown that the alignment between the gradients of the optical flow components and between the latter and the intensity gradients is an important measure of the flow’s quality. Adding this criterion as a requirement in the optimization process improves the resulting flow. This is demonstrated in synthetic and real sequences.