In recent years, there has been a proliferation of spatiotemporal foundation models in different scientific disciplines. While promising, these models are often domain-specific and are only assessed within the particular applications for which they are designed. Given that many tasks can be represented as video modeling problems, video foundation models (ViFMs) hold considerable promise as general-purpose domain-agnostic approaches. However, it is not known whether the knowledge acquired on large-scale but potentially out-of-domain data can be effectively transferred across diverse scientific disciplines, and if a single, pretrained ViFM can be competitive with domain-specific baselines. To address this, we introduce SciVid, a comprehensive benchmark comprising five *Sci*entific *Vid*eo tasks, across medical computer vision, animal behavior, and weather forecasting. We adapt six leading ViFMs to SciVid using simple trainable readout modules, establishing strong baselines and demonstrating the potential for effective transfer learning. Specifically, we show that state-of-the-art results can be obtained in several applications by leveraging the general-purpose representations from ViFM backbones. Furthermore, our results reveal the limitations of existing ViFMs, and highlight opportunities for the development of generalizable models for high-impact scientific applications. We release our code at https://github.com/google-deepmind/scivid to facilitate further research in the development of ViFMs.

Generic Object Tracking (GOT) is the problem of tracking target objects, specified by bounding boxes in the first frame of a video. While the task has received much attention in the last decades, researchers have almost exclusively focused on the single object setting. However multiobject GOT poses its own challenges and is more attractive in real-world applications. We attribute the lack of research interest into this problem to the absence of suitable benchmarks. In this work, we introduce a new largescale GOT benchmark, LaGOT, containing multiple annotated target objects per sequence. Our benchmark allows users to tackle key remaining challenges in GOT, aiming to increase robustness and reduce computation through joint tracking of multiple objects simultaneously. In addition, we propose a transformer-based GOT tracker baseline capable of joint processing of multiple objects through shared computation. Our approach achieves a 4× faster run-time in case of 10 concurrent objects compared to tracking each object independently and outperforms existing single object trackers on our new benchmark. In addition, our approach achieves highly competitive results on single-object GOT datasets, setting a new state of the art on TrackingNet with a success rate AUC of 84.4%. Our benchmark, code, results and trained models are available at https://github.com/visionml/pytracking.

Transfer learning enables to re-use knowledge learned on a source task to help learning a target task. A simple form of transfer learning is common in current state-of-the-art computer vision models, i.e., pre-training a model for image classification on the ILSVRC dataset, and then fine-tune on any target task. However, previous systematic studies of transfer learning have been limited and the circumstances in which it is expected to work are not fully understood. In this paper we carry out an extensive experimental exploration of transfer learning across vastly different image domains (consumer photos, autonomous driving, aerial imagery, underwater, indoor scenes, synthetic, close-ups) and task types (semantic segmentation, object detection, depth estimation, keypoint detection). Importantly, these are all complex, structured output tasks types relevant to modern computer vision applications. In total we carry out over 2000 transfer learning experiments, including many where the source and target come from different image domains, task types, or both. We systematically analyze these experiments to understand the impact of image domain, task type, and dataset size on transfer learning performance. Our study leads to several insights and concrete recommendations: (1) for most tasks there exists a source which significantly outperforms ILSVRC'12 pre-training; (2) the image domain is the most important factor for achieving positive transfer; (3) the source dataset should include the image domain of the target dataset to achieve best results; (4) at the same time, we observe only small negative effects when the image domain of the source task is much broader than that of the target; (5) transfer across task types can be beneficial, but its success is heavily dependent on both the source and target task types.

We introduce a unified framework for generic video annotation with bounding boxes. Video annotation is a longstanding problem, as it is a tedious and time-consuming process. We tackle two important challenges of video annotation: (1) automatic temporal interpolation and extrapolation of bounding boxes provided by a human annotator on a subset of all frames, and (2) automatic selection of frames to annotate manually. Our contribution is two-fold: first, we propose a model that has both interpolating and extrapolating capabilities; second, we propose a guiding mechanism that sequentially generates suggestions for what frame to annotate next, based on the annotations made previously. We extensively evaluate our approach on several challenging datasets in simulation and demonstrate a reduction in terms of the number of manual bounding boxes drawn by 60% over linear interpolation and by 35% over an off-the-shelf tracker. Moreover, we also show 10% annotation time improvement over a state-of-the-art method for video annotation with bounding boxes [25]. Finally, we run human annotation experiments and provide extensive analysis of the results, showing that our approach reduces actual measured annotation time by 50% compared to commonly used linear interpolation.

We present Open Images V4, a dataset of 9.2M images with unified annotations for image classification, object detection and visual relationship detection. The images have a Creative Commons Attribution license that allows to share and adapt the material, and they have been collected from Flickr without a predefined list of class names or tags, leading to natural class statistics and avoiding an initial design bias. Open Images V4 offers large scale across several dimensions: 30.1M image-level labels for 19.8k concepts, 15.4M bounding boxes for 600 object classes, and 375k visual relationship annotations involving 57 classes. For object detection in particular, we provide 15x more bounding boxes than the next largest datasets (15.4M boxes on 1.9M images). The images often show complex scenes with several objects (8 annotated objects per image on average). We annotated visual relationships between them, which support visual relationship detection, an emerging task that requires structured reasoning. We provide in-depth comprehensive statistics about the dataset, we validate the quality of the annotations, we study how the performance of several modern models evolves with increasing amounts of training data, and we demonstrate two applications made possible by having unified annotations of multiple types coexisting in the same images. We hope that the scale, quality, and variety of Open Images V4 will foster further research and innovation even beyond the areas of image classification, object detection, and visual relationship detection.

Viewpoint estimation, especially in case of multiple object classes, remains an important and challenging problem. First, objects under different views undergo extreme appearance variations, often making within-class variance larger than between-class variance. Second, obtaining precise ground truth for real-world images, necessary for training supervised viewpoint estimation models, is extremely difficult and time consuming. As a result, annotated data is often available only for a limited number of classes. Hence it is desirable to share viewpoint information across classes. Additional complexity arises from unaligned pose labels between classes, i.e. a side view of a car might look more like a frontal view of a toaster, than its side view. To address these problems, we propose a metric learning approach for joint class prediction and pose estimation. Our approach allows to circumvent the problem of viewpoint alignment across multiple classes, and does not require dense viewpoint labels. Moreover, we show, that the learned metric generalizes to new classes, for which the pose labels are not available, and therefore makes it possible to use only partially annotated training sets, relying on the intrinsic similarities in the viewpoint manifolds. We evaluate our approach on two challenging multi-class datasets, 3DObjects and PASCAL3D+.

Over the last several years it has been shown that image-based object detectors are sensitive to the training data and often fail to generalize to examples that fall outside the original training sample domain (e.g., videos). A number of domain adaptation (DA) techniques have been proposed to address this problem. DA approaches are designed to adapt a fixed complexity model to the new (e.g., video) domain. We posit that unlabeled data should not only allow adaptation, but also improve (or at least maintain) performance on the original and other domains by dynamically adjusting model complexity and parameters. We call this notion domain expansion. To this end, we develop a new scalable and accurate incremental object detection algorithm, based on several extensions of large-margin embedding (LME). Our detection model consists of an embedding space and multiple class prototypes in that embedding space, that represent object classes; distance to those prototypes allows us to reason about multi-class detection. By incrementally detecting object instances in video and adding confident detections into the model, we are able to dynamically adjust the complexity of the detector over time by instantiating new prototypes to span all domains the model has seen. We test performance of our approach by expanding an object detector trained on ImageNet to detect objects in egocentric videos of Activity Daily Living (ADL) dataset and challenging videos from YouTube Objects (YTO) dataset.

We present a novel method for multiple people tracking that leverages a generalized model for capturing interactions among individuals. At the core of our model lies a learned dictionary of interaction feature strings which capture relationships between the motions of targets. These feature strings, created from low-level image features, lead to a much richer representation of the physical interactions between targets compared to hand-specified social force models that previous works have introduced for tracking. One disadvantage of using social forces is that all pedestrians must be detected in order for the forces to be applied, while our method is able to encode the effect of undetected targets, making the tracker more robust to partial occlusions. The interaction feature strings are used in a Random Forest framework to track targets according to the features surrounding them. Results on six publicly available sequences show that our method outperforms state-of-the-art approaches in multiple people tracking.