Alexander Kirillov

Aaron Hurst, Adam Lerer, Adam P. Goucher et al. · openai

GPT-4o is an autoregressive omni model that accepts as input any combination of text, audio, image, and video, and generates any combination of text, audio, and image outputs. It's trained end-to-end across text, vision, and audio, meaning all inputs and outputs are processed by the same neural network. GPT-4o can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds, which is similar to human response time in conversation. It matches GPT-4 Turbo performance on text in English and code, with significant improvement on text in non-English languages, while also being much faster and 50\% cheaper in the API. GPT-4o is especially better at vision and audio understanding compared to existing models. In line with our commitment to building AI safely and consistent with our voluntary commitments to the White House, we are sharing the GPT-4o System Card, which includes our Preparedness Framework evaluations. In this System Card, we provide a detailed look at GPT-4o's capabilities, limitations, and safety evaluations across multiple categories, focusing on speech-to-speech while also evaluating text and image capabilities, and measures we've implemented to ensure the model is safe and aligned. We also include third-party assessments on dangerous capabilities, as well as discussion of potential societal impacts of GPT-4o's text and vision capabilities.

Duy-Kien Nguyen, Vaibhav Aggarwal, Yanghao Li et al. · meta-ai

In this work, we explore regions as a potential visual analogue of words for self-supervised image representation learning. Inspired by Masked Autoencoding (MAE), a generative pre-training baseline, we propose masked region autoencoding to learn from groups of pixels or regions. Specifically, we design an architecture which efficiently addresses the one-to-many mapping between images and regions, while being highly effective especially with high-quality regions. When integrated with MAE, our approach (R-MAE) demonstrates consistent improvements across various pre-training datasets and downstream detection and segmentation benchmarks, with negligible computational overheads. Beyond the quantitative evaluation, our analysis indicates the models pre-trained with masked region autoencoding unlock the potential for interactive segmentation. The code is provided at https://github.com/facebookresearch/r-mae.

Alexander Kirillov, Eric Mintun, Nikhila Ravi et al. · meta-ai

We introduce the Segment Anything (SA) project: a new task, model, and dataset for image segmentation. Using our efficient model in a data collection loop, we built the largest segmentation dataset to date (by far), with over 1 billion masks on 11M licensed and privacy respecting images. The model is designed and trained to be promptable, so it can transfer zero-shot to new image distributions and tasks. We evaluate its capabilities on numerous tasks and find that its zero-shot performance is impressive -- often competitive with or even superior to prior fully supervised results. We are releasing the Segment Anything Model (SAM) and corresponding dataset (SA-1B) of 1B masks and 11M images at https://segment-anything.com to foster research into foundation models for computer vision.

Eric Mintun, Alexander Kirillov, Saining Xie

Invariance to a broad array of image corruptions, such as warping, noise, or color shifts, is an important aspect of building robust models in computer vision. Recently, several new data augmentations have been proposed that significantly improve performance on ImageNet-C, a benchmark of such corruptions. However, there is still a lack of basic understanding on the relationship between data augmentations and test-time corruptions. To this end, we develop a feature space for image transforms, and then use a new measure in this space between augmentations and corruptions called the Minimal Sample Distance to demonstrate a strong correlation between similarity and performance. We then investigate recent data augmentations and observe a significant degradation in corruption robustness when the test-time corruptions are sampled to be perceptually dissimilar from ImageNet-C in this feature space. Our results suggest that test error can be improved by training on perceptually similar augmentations, and data augmentations may not generalize well beyond the existing benchmark. We hope our results and tools will allow for more robust progress towards improving robustness to image corruptions. We provide code at https://github.com/facebookresearch/augmentation-corruption.

Tim Meinhardt, Alexander Kirillov, Laura Leal-Taixe et al.

The challenging task of multi-object tracking (MOT) requires simultaneous reasoning about track initialization, identity, and spatio-temporal trajectories. We formulate this task as a frame-to-frame set prediction problem and introduce TrackFormer, an end-to-end trainable MOT approach based on an encoder-decoder Transformer architecture. Our model achieves data association between frames via attention by evolving a set of track predictions through a video sequence. The Transformer decoder initializes new tracks from static object queries and autoregressively follows existing tracks in space and time with the conceptually new and identity preserving track queries. Both query types benefit from self- and encoder-decoder attention on global frame-level features, thereby omitting any additional graph optimization or modeling of motion and/or appearance. TrackFormer introduces a new tracking-by-attention paradigm and while simple in its design is able to achieve state-of-the-art performance on the task of multi-object tracking (MOT17 and MOT20) and segmentation (MOTS20). The code is available at https://github.com/timmeinhardt/trackformer .

Nicolas Carion, Francisco Massa, Gabriel Synnaeve et al.

We present a new method that views object detection as a direct set prediction problem. Our approach streamlines the detection pipeline, effectively removing the need for many hand-designed components like a non-maximum suppression procedure or anchor generation that explicitly encode our prior knowledge about the task. The main ingredients of the new framework, called DEtection TRansformer or DETR, are a set-based global loss that forces unique predictions via bipartite matching, and a transformer encoder-decoder architecture. Given a fixed small set of learned object queries, DETR reasons about the relations of the objects and the global image context to directly output the final set of predictions in parallel. The new model is conceptually simple and does not require a specialized library, unlike many other modern detectors. DETR demonstrates accuracy and run-time performance on par with the well-established and highly-optimized Faster RCNN baseline on the challenging COCO object detection dataset. Moreover, DETR can be easily generalized to produce panoptic segmentation in a unified manner. We show that it significantly outperforms competitive baselines. Training code and pretrained models are available at https://github.com/facebookresearch/detr.

Alexander Kirillov, Yuxin Wu, Kaiming He et al.

We present a new method for efficient high-quality image segmentation of objects and scenes. By analogizing classical computer graphics methods for efficient rendering with over- and undersampling challenges faced in pixel labeling tasks, we develop a unique perspective of image segmentation as a rendering problem. From this vantage, we present the PointRend (Point-based Rendering) neural network module: a module that performs point-based segmentation predictions at adaptively selected locations based on an iterative subdivision algorithm. PointRend can be flexibly applied to both instance and semantic segmentation tasks by building on top of existing state-of-the-art models. While many concrete implementations of the general idea are possible, we show that a simple design already achieves excellent results. Qualitatively, PointRend outputs crisp object boundaries in regions that are over-smoothed by previous methods. Quantitatively, PointRend yields significant gains on COCO and Cityscapes, for both instance and semantic segmentation. PointRend's efficiency enables output resolutions that are otherwise impractical in terms of memory or computation compared to existing approaches. Code has been made available at https://github.com/facebookresearch/detectron2/tree/master/projects/PointRend.

Yutong Bai, Xinlei Chen, Alexander Kirillov et al.

In this work we present point-level region contrast, a self-supervised pre-training approach for the task of object detection. This approach is motivated by the two key factors in detection: localization and recognition. While accurate localization favors models that operate at the pixel- or point-level, correct recognition typically relies on a more holistic, region-level view of objects. Incorporating this perspective in pre-training, our approach performs contrastive learning by directly sampling individual point pairs from different regions. Compared to an aggregated representation per region, our approach is more robust to the change in input region quality, and further enables us to implicitly improve initial region assignments via online knowledge distillation during training. Both advantages are important when dealing with imperfect regions encountered in the unsupervised setting. Experiments show point-level region contrast improves on state-of-the-art pre-training methods for object detection and segmentation across multiple tasks and datasets, and we provide extensive ablation studies and visualizations to aid understanding. Code will be made available.

Norman Mu, Alexander Kirillov, David Wagner et al.

Recent work has shown that self-supervised pre-training leads to improvements over supervised learning on challenging visual recognition tasks. CLIP, an exciting new approach to learning with language supervision, demonstrates promising performance on a wide variety of benchmarks. In this work, we explore whether self-supervised learning can aid in the use of language supervision for visual representation learning. We introduce SLIP, a multi-task learning framework for combining self-supervised learning and CLIP pre-training. After pre-training with Vision Transformers, we thoroughly evaluate representation quality and compare performance to both CLIP and self-supervised learning under three distinct settings: zero-shot transfer, linear classification, and end-to-end finetuning. Across ImageNet and a battery of additional datasets, we find that SLIP improves accuracy by a large margin. We validate our results further with experiments on different model sizes, training schedules, and pre-training datasets. Our findings show that SLIP enjoys the best of both worlds: better performance than self-supervision (+8.1% linear accuracy) and language supervision (+5.2% zero-shot accuracy).

Bowen Cheng, Anwesa Choudhuri, Ishan Misra et al.

We find Mask2Former also achieves state-of-the-art performance on video instance segmentation without modifying the architecture, the loss or even the training pipeline. In this report, we show universal image segmentation architectures trivially generalize to video segmentation by directly predicting 3D segmentation volumes. Specifically, Mask2Former sets a new state-of-the-art of 60.4 AP on YouTubeVIS-2019 and 52.6 AP on YouTubeVIS-2021. We believe Mask2Former is also capable of handling video semantic and panoptic segmentation, given its versatility in image segmentation. We hope this will make state-of-the-art video segmentation research more accessible and bring more attention to designing universal image and video segmentation architectures.

Bowen Cheng, Ishan Misra, Alexander G. Schwing et al.

Image segmentation is about grouping pixels with different semantics, e.g., category or instance membership, where each choice of semantics defines a task. While only the semantics of each task differ, current research focuses on designing specialized architectures for each task. We present Masked-attention Mask Transformer (Mask2Former), a new architecture capable of addressing any image segmentation task (panoptic, instance or semantic). Its key components include masked attention, which extracts localized features by constraining cross-attention within predicted mask regions. In addition to reducing the research effort by at least three times, it outperforms the best specialized architectures by a significant margin on four popular datasets. Most notably, Mask2Former sets a new state-of-the-art for panoptic segmentation (57.8 PQ on COCO), instance segmentation (50.1 AP on COCO) and semantic segmentation (57.7 mIoU on ADE20K).

Shengyi Qian, Alexander Kirillov, Nikhila Ravi et al.

Humans can perceive scenes in 3D from a handful of 2D views. For AI agents, the ability to recognize a scene from any viewpoint given only a few images enables them to efficiently interact with the scene and its objects. In this work, we attempt to endow machines with this ability. We propose a model which takes as input a few RGB images of a new scene and recognizes the scene from novel viewpoints by segmenting it into semantic categories. All this without access to the RGB images from those views. We pair 2D scene recognition with an implicit 3D representation and learn from multi-view 2D annotations of hundreds of scenes without any 3D supervision beyond camera poses. We experiment on challenging datasets and demonstrate our model's ability to jointly capture semantics and geometry of novel scenes with diverse layouts, object types and shapes.

Bowen Cheng, Alexander G. Schwing, Alexander Kirillov

Modern approaches typically formulate semantic segmentation as a per-pixel classification task, while instance-level segmentation is handled with an alternative mask classification. Our key insight: mask classification is sufficiently general to solve both semantic- and instance-level segmentation tasks in a unified manner using the exact same model, loss, and training procedure. Following this observation, we propose MaskFormer, a simple mask classification model which predicts a set of binary masks, each associated with a single global class label prediction. Overall, the proposed mask classification-based method simplifies the landscape of effective approaches to semantic and panoptic segmentation tasks and shows excellent empirical results. In particular, we observe that MaskFormer outperforms per-pixel classification baselines when the number of classes is large. Our mask classification-based method outperforms both current state-of-the-art semantic (55.6 mIoU on ADE20K) and panoptic segmentation (52.7 PQ on COCO) models.

Bowen Cheng, Omkar Parkhi, Alexander Kirillov

We propose an embarrassingly simple point annotation scheme to collect weak supervision for instance segmentation. In addition to bounding boxes, we collect binary labels for a set of points uniformly sampled inside each bounding box. We show that the existing instance segmentation models developed for full mask supervision can be seamlessly trained with point-based supervision collected via our scheme. Remarkably, Mask R-CNN trained on COCO, PASCAL VOC, Cityscapes, and LVIS with only 10 annotated random points per object achieves 94%--98% of its fully-supervised performance, setting a strong baseline for weakly-supervised instance segmentation. The new point annotation scheme is approximately 5 times faster than annotating full object masks, making high-quality instance segmentation more accessible in practice. Inspired by the point-based annotation form, we propose a modification to PointRend instance segmentation module. For each object, the new architecture, called Implicit PointRend, generates parameters for a function that makes the final point-level mask prediction. Implicit PointRend is more straightforward and uses a single point-level mask loss. Our experiments show that the new module is more suitable for the point-based supervision.

Bowen Cheng, Ross Girshick, Piotr Dollár et al.

We present Boundary IoU (Intersection-over-Union), a new segmentation evaluation measure focused on boundary quality. We perform an extensive analysis across different error types and object sizes and show that Boundary IoU is significantly more sensitive than the standard Mask IoU measure to boundary errors for large objects and does not over-penalize errors on smaller objects. The new quality measure displays several desirable characteristics like symmetry w.r.t. prediction/ground truth pairs and balanced responsiveness across scales, which makes it more suitable for segmentation evaluation than other boundary-focused measures like Trimap IoU and F-measure. Based on Boundary IoU, we update the standard evaluation protocols for instance and panoptic segmentation tasks by proposing the Boundary AP (Average Precision) and Boundary PQ (Panoptic Quality) metrics, respectively. Our experiments show that the new evaluation metrics track boundary quality improvements that are generally overlooked by current Mask IoU-based evaluation metrics. We hope that the adoption of the new boundary-sensitive evaluation metrics will lead to rapid progress in segmentation methods that improve boundary quality.

Achal Dave, Piotr Dollár, Deva Ramanan et al.

By design, average precision (AP) for object detection aims to treat all classes independently: AP is computed independently per category and averaged. On one hand, this is desirable as it treats all classes equally. On the other hand, it ignores cross-category confidence calibration, a key property in real-world use cases. Unfortunately, under important conditions (i.e., large vocabulary, high instance counts) the default implementation of AP is neither category independent, nor does it directly reward properly calibrated detectors. In fact, we show that on LVIS the default implementation produces a gameable metric, where a simple, un-intuitive re-ranking policy can improve AP by a large margin. To address these limitations, we introduce two complementary metrics. First, we present a simple fix to the default AP implementation, ensuring that it is independent across categories as originally intended. We benchmark recent LVIS detection advances and find that many reported gains do not translate to improvements under our new evaluation, suggesting recent improvements may arise from difficult to interpret changes to cross-category rankings. Given the importance of reliably benchmarking cross-category rankings, we consider a pooled version of AP (AP-Pool) that rewards properly calibrated detectors by directly comparing cross-category rankings. Finally, we revisit classical approaches for calibration and find that explicitly calibrating detectors improves state-of-the-art on AP-Pool by 1.7 points

Saining Xie, Alexander Kirillov, Ross Girshick et al.

Neural networks for image recognition have evolved through extensive manual design from simple chain-like models to structures with multiple wiring paths. The success of ResNets and DenseNets is due in large part to their innovative wiring plans. Now, neural architecture search (NAS) studies are exploring the joint optimization of wiring and operation types, however, the space of possible wirings is constrained and still driven by manual design despite being searched. In this paper, we explore a more diverse set of connectivity patterns through the lens of randomly wired neural networks. To do this, we first define the concept of a stochastic network generator that encapsulates the entire network generation process. Encapsulation provides a unified view of NAS and randomly wired networks. Then, we use three classical random graph models to generate randomly wired graphs for networks. The results are surprising: several variants of these random generators yield network instances that have competitive accuracy on the ImageNet benchmark. These results suggest that new efforts focusing on designing better network generators may lead to new breakthroughs by exploring less constrained search spaces with more room for novel design.

Alexander Kirillov, Ross Girshick, Kaiming He et al.

The recently introduced panoptic segmentation task has renewed our community's interest in unifying the tasks of instance segmentation (for thing classes) and semantic segmentation (for stuff classes). However, current state-of-the-art methods for this joint task use separate and dissimilar networks for instance and semantic segmentation, without performing any shared computation. In this work, we aim to unify these methods at the architectural level, designing a single network for both tasks. Our approach is to endow Mask R-CNN, a popular instance segmentation method, with a semantic segmentation branch using a shared Feature Pyramid Network (FPN) backbone. Surprisingly, this simple baseline not only remains effective for instance segmentation, but also yields a lightweight, top-performing method for semantic segmentation. In this work, we perform a detailed study of this minimally extended version of Mask R-CNN with FPN, which we refer to as Panoptic FPN, and show it is a robust and accurate baseline for both tasks. Given its effectiveness and conceptual simplicity, we hope our method can serve as a strong baseline and aid future research in panoptic segmentation.

Alexander Kirillov, Natalia Krizhanovsky, Andrew Krizhanovsky

The problem of word sense disambiguation (WSD) is considered in the article. Given a set of synonyms (synsets) and sentences with these synonyms. It is necessary to select the meaning of the word in the sentence automatically. 1285 sentences were tagged by experts, namely, one of the dictionary meanings was selected by experts for target words. To solve the WSD-problem, an algorithm based on a new method of vector-word contexts proximity calculation is proposed. In order to achieve higher accuracy, a preliminary epsilon-filtering of words is performed, both in the sentence and in the set of synonyms. An extensive program of experiments was carried out. Four algorithms are implemented, including a new algorithm. Experiments have shown that in a number of cases the new algorithm shows better results. The developed software and the tagged corpus have an open license and are available online. Wiktionary and Wikisource are used. A brief description of this work can be viewed in slides (https://goo.gl/9ak6Gt). Video lecture in Russian on this research is available online (https://youtu.be/-DLmRkepf58).

Andrew Krizhanovsky, Alexander Kirillov

The goal of formalization, proposed in this paper, is to bring together, as near as possible, the theoretic linguistic problem of synonym conception and the computer linguistic methods based generally on empirical intuitive unjustified factors. Using the word vector representation we have proposed the geometric approach to mathematical modeling of synonym set (synset). The word embedding is based on the neural networks (Skip-gram, CBOW), developed and realized as word2vec program by T. Mikolov. The standard cosine similarity is used as the distance between word-vectors. Several geometric characteristics of the synset words are introduced: the interior of synset, the synset word rank and centrality. These notions are intended to select the most significant synset words, i.e. the words which senses are the nearest to the sense of a synset. Some experiments with proposed notions, based on RusVectores resources, are represented. A brief description of this work can be viewed in slides https://goo.gl/K82Fei

Alexander Kirillov, Kaiming He, Ross Girshick et al.

We propose and study a task we name panoptic segmentation (PS). Panoptic segmentation unifies the typically distinct tasks of semantic segmentation (assign a class label to each pixel) and instance segmentation (detect and segment each object instance). The proposed task requires generating a coherent scene segmentation that is rich and complete, an important step toward real-world vision systems. While early work in computer vision addressed related image/scene parsing tasks, these are not currently popular, possibly due to lack of appropriate metrics or associated recognition challenges. To address this, we propose a novel panoptic quality (PQ) metric that captures performance for all classes (stuff and things) in an interpretable and unified manner. Using the proposed metric, we perform a rigorous study of both human and machine performance for PS on three existing datasets, revealing interesting insights about the task. The aim of our work is to revive the interest of the community in a more unified view of image segmentation.

Omid Hosseini Jafari, Oliver Groth, Alexander Kirillov et al.

This paper addresses the task of designing a modular neural network architecture that jointly solves different tasks. As an example we use the tasks of depth estimation and semantic segmentation given a single RGB image. The main focus of this work is to analyze the cross-modality influence between depth and semantic prediction maps on their joint refinement. While most previous works solely focus on measuring improvements in accuracy, we propose a way to quantify the cross-modality influence. We show that there is a relationship between final accuracy and cross-modality influence, although not a simple linear one. Hence a larger cross-modality influence does not necessarily translate into an improved accuracy. We find that a beneficial balance between the cross-modality influences can be achieved by network architecture and conjecture that this relationship can be utilized to understand different network design choices. Towards this end we propose a Convolutional Neural Network (CNN) architecture that fuses the state of the state-of-the-art results for depth estimation and semantic labeling. By balancing the cross-modality influences between depth and semantic prediction, we achieve improved results for both tasks using the NYU-Depth v2 benchmark.

Frank Michel, Alexander Kirillov, Eric Brachmann et al.

This paper addresses the task of estimating the 6D pose of a known 3D object from a single RGB-D image. Most modern approaches solve this task in three steps: i) Compute local features; ii) Generate a pool of pose-hypotheses; iii) Select and refine a pose from the pool. This work focuses on the second step. While all existing approaches generate the hypotheses pool via local reasoning, e.g. RANSAC or Hough-voting, we are the first to show that global reasoning is beneficial at this stage. In particular, we formulate a novel fully-connected Conditional Random Field (CRF) that outputs a very small number of pose-hypotheses. Despite the potential functions of the CRF being non-Gaussian, we give a new and efficient two-step optimization procedure, with some guarantees for optimality. We utilize our global hypotheses generation procedure to produce results that exceed state-of-the-art for the challenging "Occluded Object Dataset".

Alexander Kirillov, Evgeny Levinkov, Bjoern Andres et al.

This work addresses the task of instance-aware semantic segmentation. Our key motivation is to design a simple method with a new modelling-paradigm, which therefore has a different trade-off between advantages and disadvantages compared to known approaches. Our approach, we term InstanceCut, represents the problem by two output modalities: (i) an instance-agnostic semantic segmentation and (ii) all instance-boundaries. The former is computed from a standard convolutional neural network for semantic segmentation, and the latter is derived from a new instance-aware edge detection model. To reason globally about the optimal partitioning of an image into instances, we combine these two modalities into a novel MultiCut formulation. We evaluate our approach on the challenging CityScapes dataset. Despite the conceptual simplicity of our approach, we achieve the best result among all published methods, and perform particularly well for rare object classes.

Evgeny Levinkov, Jonas Uhrig, Siyu Tang et al.

We state a combinatorial optimization problem whose feasible solutions define both a decomposition and a node labeling of a given graph. This problem offers a common mathematical abstraction of seemingly unrelated computer vision tasks, including instance-separating semantic segmentation, articulated human body pose estimation and multiple object tracking. Conceptually, the problem we state generalizes the unconstrained integer quadratic program and the minimum cost lifted multicut problem, both of which are NP-hard. In order to find feasible solutions efficiently, we define two local search algorithms that converge monotonously to a local optimum, offering a feasible solution at any time. To demonstrate their effectiveness in tackling computer vision tasks, we apply these algorithms to instances of the problem that we construct from published data, using published algorithms. We report state-of-the-art application-specific accuracy for the three above-mentioned applications.

Alexander Kirillov, Alexander Shekhovtsov, Carsten Rother et al.

We consider the problem of jointly inferring the M-best diverse labelings for a binary (high-order) submodular energy of a graphical model. Recently, it was shown that this problem can be solved to a global optimum, for many practically interesting diversity measures. It was noted that the labelings are, so-called, nested. This nestedness property also holds for labelings of a class of parametric submodular minimization problems, where different values of the global parameter $γ$ give rise to different solutions. The popular example of the parametric submodular minimization is the monotonic parametric max-flow problem, which is also widely used for computing multiple labelings. As the main contribution of this work we establish a close relationship between diversity with submodular energies and the parametric submodular minimization. In particular, the joint M-best diverse labelings can be obtained by running a non-parametric submodular minimization (in the special case - max-flow) solver for M different values of $γ$ in parallel, for certain diversity measures. Importantly, the values for $γ$ can be computed in a closed form in advance, prior to any optimization. These theoretical results suggest two simple yet efficient algorithms for the joint M-best diverse problem, which outperform competitors in terms of runtime and quality of results. In particular, as we show in the paper, the new methods compute the exact M-best diverse labelings faster than a popular method of Batra et al., which in some sense only obtains approximate solutions.

Alexander Kirillov, Dmitrij Schlesinger, Shuai Zheng et al.

We propose a new CNN-CRF end-to-end learning framework, which is based on joint stochastic optimization with respect to both Convolutional Neural Network (CNN) and Conditional Random Field (CRF) parameters. While stochastic gradient descent is a standard technique for CNN training, it was not used for joint models so far. We show that our learning method is (i) general, i.e. it applies to arbitrary CNN and CRF architectures and potential functions; (ii) scalable, i.e. it has a low memory footprint and straightforwardly parallelizes on GPUs; (iii) easy in implementation. Additionally, the unified CNN-CRF optimization approach simplifies a potential hardware implementation. We empirically evaluate our method on the task of semantic labeling of body parts in depth images and show that it compares favorably to competing techniques.