Andrew Rabinovich

Yimeng Bai, Chang Liu, Yang Zhang et al.

Generative recommendation is emerging as a transformative paradigm by directly generating recommended items, rather than relying on matching. Building such a system typically involves two key components: (1) optimizing the tokenizer to derive suitable item identifiers, and (2) training the recommender based on those identifiers. Existing approaches often treat these components separately--either sequentially or in alternation--overlooking their interdependence. This separation can lead to misalignment: the tokenizer is trained without direct guidance from the recommendation objective, potentially yielding suboptimal identifiers that degrade recommendation performance. To address this, we propose BLOGER, a Bi-Level Optimization for GEnerative Recommendation framework, which explicitly models the interdependence between the tokenizer and the recommender in a unified optimization process. The lower level trains the recommender using tokenized sequences, while the upper level optimizes the tokenizer based on both the tokenization loss and recommendation loss. We adopt a meta-learning approach to solve this bi-level optimization efficiently, and introduce gradient surgery to mitigate gradient conflicts in the upper-level updates, thereby ensuring that item identifiers are both informative and recommendation-aligned. Extensive experiments on multiple real-world datasets demonstrate that BLOGER consistently outperforms state-of-the-art generative recommendation methods while maintaining practical efficiency with no significant additional computational overhead, effectively bridging the gap between item tokenization and autoregressive generation. We release our code at https://github.com/Ten-Mao/BLOGER.

Ayan Sinha, Zak Murez, James Bartolozzi et al.

Multi-view stereo (MVS) is the golden mean between the accuracy of active depth sensing and the practicality of monocular depth estimation. Cost volume based approaches employing 3D convolutional neural networks (CNNs) have considerably improved the accuracy of MVS systems. However, this accuracy comes at a high computational cost which impedes practical adoption. Distinct from cost volume approaches, we propose an efficient depth estimation approach by first (a) detecting and evaluating descriptors for interest points, then (b) learning to match and triangulate a small set of interest points, and finally (c) densifying this sparse set of 3D points using CNNs. An end-to-end network efficiently performs all three steps within a deep learning framework and trained with intermediate 2D image and 3D geometric supervision, along with depth supervision. Crucially, our first step complements pose estimation using interest point detection and descriptor learning. We demonstrate state-of-the-art results on depth estimation with lower compute for different scene lengths. Furthermore, our method generalizes to newer environments and the descriptors output by our network compare favorably to strong baselines. Code is available at https://github.com/magicleap/DELTAS

Paul-Edouard Sarlin, Daniel DeTone, Tomasz Malisiewicz et al.

This paper introduces SuperGlue, a neural network that matches two sets of local features by jointly finding correspondences and rejecting non-matchable points. Assignments are estimated by solving a differentiable optimal transport problem, whose costs are predicted by a graph neural network. We introduce a flexible context aggregation mechanism based on attention, enabling SuperGlue to reason about the underlying 3D scene and feature assignments jointly. Compared to traditional, hand-designed heuristics, our technique learns priors over geometric transformations and regularities of the 3D world through end-to-end training from image pairs. SuperGlue outperforms other learned approaches and achieves state-of-the-art results on the task of pose estimation in challenging real-world indoor and outdoor environments. The proposed method performs matching in real-time on a modern GPU and can be readily integrated into modern SfM or SLAM systems. The code and trained weights are publicly available at https://github.com/magicleap/SuperGluePretrainedNetwork.

Wei Liu, Andrew Rabinovich, Alexander C. Berg

We present a technique for adding global context to deep convolutional networks for semantic segmentation. The approach is simple, using the average feature for a layer to augment the features at each location. In addition, we study several idiosyncrasies of training, significantly increasing the performance of baseline networks (e.g. from FCN). When we add our proposed global feature, and a technique for learning normalization parameters, accuracy increases consistently even over our improved versions of the baselines. Our proposed approach, ParseNet, achieves state-of-the-art performance on SiftFlow and PASCAL-Context with small additional computational cost over baselines, and near current state-of-the-art performance on PASCAL VOC 2012 semantic segmentation with a simple approach. Code is available at https://github.com/weiliu89/caffe/tree/fcn .

Mikołaj Sacha, Hammad Jafri, Mattie Terzolo et al.

Recommending matches in a text-rich, dynamic two-sided marketplace presents unique challenges due to evolving content and interaction graphs. We introduce GraphMatch, a new large-scale recommendation framework that fuses pre-trained language models with graph neural networks to overcome these challenges. Unlike prior approaches centered on standalone models, GraphMatch is a comprehensive recipe built on powerful text encoders and GNNs working in tandem. It employs adversarial negative sampling alongside point-in-time subgraph training to learn representations that capture both the fine-grained semantics of evolving text and the time-sensitive structure of the graph. We evaluated extensively on interaction data from Upwork, a leading labor marketplace, at large scale, and discuss our approach towards low-latency inference suitable for real-time use. In our experiments, GraphMatch outperforms language-only and graph-only baselines on matching tasks while being efficient at runtime. These results demonstrate that unifying language and graph representations yields a highly effective solution to text-rich, dynamic two-sided recommendations, bridging the gap between powerful pretrained LMs and large-scale graphs in practice.

Darvin Yi, Teng Liu, Mattie Terzolo et al.

As large language model (LLM) agents increasingly undertake digital work, reliable frameworks are needed to evaluate their real-world competence, adaptability, and capacity for human collaboration. Existing benchmarks remain largely static, synthetic, or domain-limited, providing limited insight into how agents perform in dynamic, economically meaningful environments. We introduce UpBench, a dynamically evolving benchmark grounded in real jobs drawn from the global Upwork labor marketplace. Each task corresponds to a verified client transaction, anchoring evaluation in genuine work activity and financial outcomes. UpBench employs a rubric-based evaluation framework, in which expert freelancers decompose each job into detailed, verifiable acceptance criteria and assess AI submissions with per-criterion feedback. This structure enables fine-grained analysis of model strengths, weaknesses, and instruction-following fidelity beyond binary pass/fail metrics. Human expertise is integrated throughout the data pipeline (from job curation and rubric construction to evaluation) ensuring fidelity to real professional standards and supporting research on human-AI collaboration. By regularly refreshing tasks to reflect the evolving nature of online work, UpBench provides a scalable, human-centered foundation for evaluating agentic systems in authentic labor-market contexts, offering a path toward a collaborative framework, where AI amplifies human capability through partnership rather than replacement.

Chenlu Ding, Jiancan Wu, Yancheng Yuan et al.

In online advertising under the cost-per-conversion (CPA) model, accurate conversion rate (CVR) prediction is crucial. A major challenge is delayed feedback, where conversions may occur long after user interactions, leading to incomplete recent data and biased model training. Existing solutions partially mitigate this issue but often rely on auxiliary models, making them computationally inefficient and less adaptive to user interest shifts. We propose IF-DFM, an \underline{I}nfluence \underline{F}unction-empowered for \underline{D}elayed \underline{F}eedback \underline{M}odeling which estimates the impact of newly arrived and delayed conversions on model parameters, enabling efficient updates without full retraining. By reformulating the inverse Hessian-vector product as an optimization problem, IF-DFM achieves a favorable trade-off between scalability and effectiveness. Experiments on benchmark datasets show that IF-DFM outperforms prior methods in both accuracy and adaptability.

Zak Murez, Tarrence van As, James Bartolozzi et al.

We present an end-to-end 3D reconstruction method for a scene by directly regressing a truncated signed distance function (TSDF) from a set of posed RGB images. Traditional approaches to 3D reconstruction rely on an intermediate representation of depth maps prior to estimating a full 3D model of a scene. We hypothesize that a direct regression to 3D is more effective. A 2D CNN extracts features from each image independently which are then back-projected and accumulated into a voxel volume using the camera intrinsics and extrinsics. After accumulation, a 3D CNN refines the accumulated features and predicts the TSDF values. Additionally, semantic segmentation of the 3D model is obtained without significant computation. This approach is evaluated on the Scannet dataset where we significantly outperform state-of-the-art baselines (deep multiview stereo followed by traditional TSDF fusion) both quantitatively and qualitatively. We compare our 3D semantic segmentation to prior methods that use a depth sensor since no previous work attempts the problem with only RGB input.

Zhengyang Wu, Srivignesh Rajendran, Tarrence van As et al.

With the emergence of Virtual and Mixed Reality (XR) devices, eye tracking has received significant attention in the computer vision community. Eye gaze estimation is a crucial component in XR -- enabling energy efficient rendering, multi-focal displays, and effective interaction with content. In head-mounted XR devices, the eyes are imaged off-axis to avoid blocking the field of view. This leads to increased challenges in inferring eye related quantities and simultaneously provides an opportunity to develop accurate and robust learning based approaches. To this end, we present MagicEyes, the first large scale eye dataset collected using real MR devices with comprehensive ground truth labeling. MagicEyes includes $587$ subjects with $80,000$ images of human-labeled ground truth and over $800,000$ images with gaze target labels. We evaluate several state-of-the-art methods on MagicEyes and also propose a new multi-task EyeNet model designed for detecting the cornea, glints and pupil along with eye segmentation in a single forward pass.

Ameya Phalak, Vijay Badrinarayanan, Andrew Rabinovich

We introduce Scan2Plan, a novel approach for accurate estimation of a floorplan from a 3D scan of the structural elements of indoor environments. The proposed method incorporates a two-stage approach where the initial stage clusters an unordered point cloud representation of the scene into room instances and wall instances using a deep neural network based voting approach. The subsequent stage estimates a closed perimeter, parameterized by a simple polygon, for each individual room by finding the shortest path along the predicted room and wall keypoints. The final floorplan is simply an assembly of all such room perimeters in the global co-ordinate system. The Scan2Plan pipeline produces accurate floorplans for complex layouts, is highly parallelizable and extremely efficient compared to existing methods. The voting module is trained only on synthetic data and evaluated on publicly available Structured3D and BKE datasets to demonstrate excellent qualitative and quantitative results outperforming state-of-the-art techniques.

Prajwal Chidananda, Ayan Sinha, Adithya Rao et al.

2D Key-point estimation is an important precursor to 3D pose estimation problems for human body and hands. In this work, we discuss the data, architecture, and training procedure necessary to deploy extremely efficient 2.5D hand pose estimation on embedded devices with highly constrained memory and compute envelope, such as AR/VR wearables. Our 2.5D hand pose estimation consists of 2D key-point estimation of joint positions on an egocentric image, captured by a depth sensor, and lifted to 2.5D using the corresponding depth values. Our contributions are two fold: (a) We discuss data labeling and augmentation strategies, the modules in the network architecture that collectively lead to $3\%$ the flop count and $2\%$ the number of parameters when compared to the state of the art MobileNetV2 architecture. (b) We propose an auxiliary multi-task training strategy needed to compensate for the small capacity of the network while achieving comparable performance to MobileNetV2. Our 32-bit trained model has a memory footprint of less than 300 Kilobytes, operates at more than 50 Hz with less than 35 MFLOPs.

Zhengyang Wu, Srivignesh Rajendran, Tarrence van As et al.

Eye gaze estimation and simultaneous semantic understanding of a user through eye images is a crucial component in Virtual and Mixed Reality; enabling energy efficient rendering, multi-focal displays and effective interaction with 3D content. In head-mounted VR/MR devices the eyes are imaged off-axis to avoid blocking the user's gaze, this view-point makes drawing eye related inferences very challenging. In this work, we present EyeNet, the first single deep neural network which solves multiple heterogeneous tasks related to eye gaze estimation and semantic user understanding for an off-axis camera setting. The tasks include eye segmentation, blink detection, emotive expression classification, IR LED glints detection, pupil and cornea center estimation. To train EyeNet end-to-end we employ both hand labelled supervision and model based supervision. We benchmark all tasks on MagicEyes, a large and new dataset of 587 subjects with varying morphology, gender, skin-color, make-up and imaging conditions.

Ameya Phalak, Zhao Chen, Darvin Yi et al.

We present DeepPerimeter, a deep learning based pipeline for inferring a full indoor perimeter (i.e. exterior boundary map) from a sequence of posed RGB images. Our method relies on robust deep methods for depth estimation and wall segmentation to generate an exterior boundary point cloud, and then uses deep unsupervised clustering to fit wall planes to obtain a final boundary map of the room. We demonstrate that DeepPerimeter results in excellent visual and quantitative performance on the popular ScanNet and FloorNet datasets and works for room shapes of various complexities as well as in multiroom scenarios. We also establish important baselines for future work on indoor perimeter estimation, topics which will become increasingly prevalent as application areas like augmented reality and robotics become more significant.

Daniel DeTone, Tomasz Malisiewicz, Andrew Rabinovich

We propose a self-supervised learning framework that uses unlabeled monocular video sequences to generate large-scale supervision for training a Visual Odometry (VO) frontend, a network which computes pointwise data associations across images. Our self-improving method enables a VO frontend to learn over time, unlike other VO and SLAM systems which require time-consuming hand-tuning or expensive data collection to adapt to new environments. Our proposed frontend operates on monocular images and consists of a single multi-task convolutional neural network which outputs 2D keypoints locations, keypoint descriptors, and a novel point stability score. We use the output of VO to create a self-supervised dataset of point correspondences to retrain the frontend. When trained using VO at scale on 2.5 million monocular images from ScanNet, the stability classifier automatically discovers a ranking for keypoints that are not likely to help in VO, such as t-junctions across depth discontinuities, features on shadows and highlights, and dynamic objects like people. The resulting frontend outperforms both traditional methods (SIFT, ORB, AKAZE) and deep learning methods (SuperPoint and LF-Net) in a 3D-to-2D pose estimation task on ScanNet.

Ayan Sinha, Zhao Chen, Vijay Badrinarayanan et al.

We propose gradient adversarial training, an auxiliary deep learning framework applicable to different machine learning problems. In gradient adversarial training, we leverage a prior belief that in many contexts, simultaneous gradient updates should be statistically indistinguishable from each other. We enforce this consistency using an auxiliary network that classifies the origin of the gradient tensor, and the main network serves as an adversary to the auxiliary network in addition to performing standard task-based training. We demonstrate gradient adversarial training for three different scenarios: (1) as a defense to adversarial examples we classify gradient tensors and tune them to be agnostic to the class of their corresponding example, (2) for knowledge distillation, we do binary classification of gradient tensors derived from the student or teacher network and tune the student gradient tensor to mimic the teacher's gradient tensor; and (3) for multi-task learning we classify the gradient tensors derived from different task loss functions and tune them to be statistically indistinguishable. For each of the three scenarios we show the potential of gradient adversarial training procedure. Specifically, gradient adversarial training increases the robustness of a network to adversarial attacks, is able to better distill the knowledge from a teacher network to a student network compared to soft targets, and boosts multi-task learning by aligning the gradient tensors derived from the task specific loss functions. Overall, our experiments demonstrate that gradient tensors contain latent information about whatever tasks are being trained, and can support diverse machine learning problems when intelligently guided through adversarialization using a auxiliary network.

Zhao Chen, Vijay Badrinarayanan, Gilad Drozdov et al.

We present a deep model that can accurately produce dense depth maps given an RGB image with known depth at a very sparse set of pixels. The model works simultaneously for both indoor/outdoor scenes and produces state-of-the-art dense depth maps at nearly real-time speeds on both the NYUv2 and KITTI datasets. We surpass the state-of-the-art for monocular depth estimation even with depth values for only 1 out of every ~10000 image pixels, and we outperform other sparse-to-dense depth methods at all sparsity levels. With depth values for 1/256 of the image pixels, we achieve a mean absolute error of less than 1% of actual depth on indoor scenes, comparable to the performance of consumer-grade depth sensor hardware. Our experiments demonstrate that it would indeed be possible to efficiently transform sparse depth measurements obtained using e.g. lower-power depth sensors or SLAM systems into high-quality dense depth maps.

Daniel DeTone, Tomasz Malisiewicz, Andrew Rabinovich

This paper presents a self-supervised framework for training interest point detectors and descriptors suitable for a large number of multiple-view geometry problems in computer vision. As opposed to patch-based neural networks, our fully-convolutional model operates on full-sized images and jointly computes pixel-level interest point locations and associated descriptors in one forward pass. We introduce Homographic Adaptation, a multi-scale, multi-homography approach for boosting interest point detection repeatability and performing cross-domain adaptation (e.g., synthetic-to-real). Our model, when trained on the MS-COCO generic image dataset using Homographic Adaptation, is able to repeatedly detect a much richer set of interest points than the initial pre-adapted deep model and any other traditional corner detector. The final system gives rise to state-of-the-art homography estimation results on HPatches when compared to LIFT, SIFT and ORB.

Zhao Chen, Vijay Badrinarayanan, Chen-Yu Lee et al.

Deep multitask networks, in which one neural network produces multiple predictive outputs, can offer better speed and performance than their single-task counterparts but are challenging to train properly. We present a gradient normalization (GradNorm) algorithm that automatically balances training in deep multitask models by dynamically tuning gradient magnitudes. We show that for various network architectures, for both regression and classification tasks, and on both synthetic and real datasets, GradNorm improves accuracy and reduces overfitting across multiple tasks when compared to single-task networks, static baselines, and other adaptive multitask loss balancing techniques. GradNorm also matches or surpasses the performance of exhaustive grid search methods, despite only involving a single asymmetry hyperparameter $α$. Thus, what was once a tedious search process that incurred exponentially more compute for each task added can now be accomplished within a few training runs, irrespective of the number of tasks. Ultimately, we will demonstrate that gradient manipulation affords us great control over the training dynamics of multitask networks and may be one of the keys to unlocking the potential of multitask learning.

Daniel DeTone, Tomasz Malisiewicz, Andrew Rabinovich

We present a point tracking system powered by two deep convolutional neural networks. The first network, MagicPoint, operates on single images and extracts salient 2D points. The extracted points are "SLAM-ready" because they are by design isolated and well-distributed throughout the image. We compare this network against classical point detectors and discover a significant performance gap in the presence of image noise. As transformation estimation is more simple when the detected points are geometrically stable, we designed a second network, MagicWarp, which operates on pairs of point images (outputs of MagicPoint), and estimates the homography that relates the inputs. This transformation engine differs from traditional approaches because it does not use local point descriptors, only point locations. Both networks are trained with simple synthetic data, alleviating the requirement of expensive external camera ground truthing and advanced graphics rendering pipelines. The system is fast and lean, easily running 30+ FPS on a single CPU.

Chen-Yu Lee, Vijay Badrinarayanan, Tomasz Malisiewicz et al.

This paper focuses on the task of room layout estimation from a monocular RGB image. Prior works break the problem into two sub-tasks: semantic segmentation of floor, walls, ceiling to produce layout hypotheses, followed by an iterative optimization step to rank these hypotheses. In contrast, we adopt a more direct formulation of this problem as one of estimating an ordered set of room layout keypoints. The room layout and the corresponding segmentation is completely specified given the locations of these ordered keypoints. We predict the locations of the room layout keypoints using RoomNet, an end-to-end trainable encoder-decoder network. On the challenging benchmark datasets Hedau and LSUN, we achieve state-of-the-art performance along with 200x to 600x speedup compared to the most recent work. Additionally, we present optional extensions to the RoomNet architecture such as including recurrent computations and memory units to refine the keypoint locations under the same parametric capacity.

Debidatta Dwibedi, Tomasz Malisiewicz, Vijay Badrinarayanan et al.

We present a Deep Cuboid Detector which takes a consumer-quality RGB image of a cluttered scene and localizes all 3D cuboids (box-like objects). Contrary to classical approaches which fit a 3D model from low-level cues like corners, edges, and vanishing points, we propose an end-to-end deep learning system to detect cuboids across many semantic categories (e.g., ovens, shipping boxes, and furniture). We localize cuboids with a 2D bounding box, and simultaneously localize the cuboid's corners, effectively producing a 3D interpretation of box-like objects. We refine keypoints by pooling convolutional features iteratively, improving the baseline method significantly. Our deep learning cuboid detector is trained in an end-to-end fashion and is suitable for real-time applications in augmented reality (AR) and robotics.

Daniel DeTone, Tomasz Malisiewicz, Andrew Rabinovich

We present a deep convolutional neural network for estimating the relative homography between a pair of images. Our feed-forward network has 10 layers, takes two stacked grayscale images as input, and produces an 8 degree of freedom homography which can be used to map the pixels from the first image to the second. We present two convolutional neural network architectures for HomographyNet: a regression network which directly estimates the real-valued homography parameters, and a classification network which produces a distribution over quantized homographies. We use a 4-point homography parameterization which maps the four corners from one image into the second image. Our networks are trained in an end-to-end fashion using warped MS-COCO images. Our approach works without the need for separate local feature detection and transformation estimation stages. Our deep models are compared to a traditional homography estimator based on ORB features and we highlight the scenarios where HomographyNet outperforms the traditional technique. We also describe a variety of applications powered by deep homography estimation, thus showcasing the flexibility of a deep learning approach.

Jonathan Malmaud, Jonathan Huang, Vivek Rathod et al.

We present a novel method for aligning a sequence of instructions to a video of someone carrying out a task. In particular, we focus on the cooking domain, where the instructions correspond to the recipe. Our technique relies on an HMM to align the recipe steps to the (automatically generated) speech transcript. We then refine this alignment using a state-of-the-art visual food detector, based on a deep convolutional neural network. We show that our technique outperforms simpler techniques based on keyword spotting. It also enables interesting applications, such as automatically illustrating recipes with keyframes, and searching within a video for events of interest.

Scott Reed, Honglak Lee, Dragomir Anguelov et al.

Current state-of-the-art deep learning systems for visual object recognition and detection use purely supervised training with regularization such as dropout to avoid overfitting. The performance depends critically on the amount of labeled examples, and in current practice the labels are assumed to be unambiguous and accurate. However, this assumption often does not hold; e.g. in recognition, class labels may be missing; in detection, objects in the image may not be localized; and in general, the labeling may be subjective. In this work we propose a generic way to handle noisy and incomplete labeling by augmenting the prediction objective with a notion of consistency. We consider a prediction consistent if the same prediction is made given similar percepts, where the notion of similarity is between deep network features computed from the input data. In experiments we demonstrate that our approach yields substantial robustness to label noise on several datasets. On MNIST handwritten digits, we show that our model is robust to label corruption. On the Toronto Face Database, we show that our model handles well the case of subjective labels in emotion recognition, achieving state-of-the- art results, and can also benefit from unlabeled face images with no modification to our method. On the ILSVRC2014 detection challenge data, we show that our approach extends to very deep networks, high resolution images and structured outputs, and results in improved scalable detection.

David Warde-Farley, Andrew Rabinovich, Dragomir Anguelov

We study the problem of large scale, multi-label visual recognition with a large number of possible classes. We propose a method for augmenting a trained neural network classifier with auxiliary capacity in a manner designed to significantly improve upon an already well-performing model, while minimally impacting its computational footprint. Using the predictions of the network itself as a descriptor for assessing visual similarity, we define a partitioning of the label space into groups of visually similar entities. We then augment the network with auxilliary hidden layer pathways with connectivity only to these groups of label units. We report a significant improvement in mean average precision on a large-scale object recognition task with the augmented model, while increasing the number of multiply-adds by less than 3%.

Christian Szegedy, Wei Liu, Yangqing Jia et al.

We propose a deep convolutional neural network architecture codenamed "Inception", which was responsible for setting the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC 2014). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. This was achieved by a carefully crafted design that allows for increasing the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC 2014 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

Samy Bengio, Jeff Dean, Dumitru Erhan et al.

Object recognition and localization are important tasks in computer vision. The focus of this work is the incorporation of contextual information in order to improve object recognition and localization. For instance, it is natural to expect not to see an elephant to appear in the middle of an ocean. We consider a simple approach to encapsulate such common sense knowledge using co-occurrence statistics from web documents. By merely counting the number of times nouns (such as elephants, sharks, oceans, etc.) co-occur in web documents, we obtain a good estimate of expected co-occurrences in visual data. We then cast the problem of combining textual co-occurrence statistics with the predictions of image-based classifiers as an optimization problem. The resulting optimization problem serves as a surrogate for our inference procedure. Albeit the simplicity of the resulting optimization problem, it is effective in improving both recognition and localization accuracy. Concretely, we observe significant improvements in recognition and localization rates for both ImageNet Detection 2012 and Sun 2012 datasets.