Amir Zamir

Yajie Bao, Yang Li, Shao-Lun Huang et al.

Task transfer learning is a popular technique in image processing applications that uses pre-trained models to reduce the supervision cost of related tasks. An important question is to determine task transferability, i.e. given a common input domain, estimating to what extent representations learned from a source task can help in learning a target task. Typically, transferability is either measured experimentally or inferred through task relatedness, which is often defined without a clear operational meaning. In this paper, we present a novel metric, H-score, an easily-computable evaluation function that estimates the performance of transferred representations from one task to another in classification problems using statistical and information theoretic principles. Experiments on real image data show that our metric is not only consistent with the empirical transferability measurement, but also useful to practitioners in applications such as source model selection and task transfer curriculum learning.

Roman Bachmann, David Mizrahi, Andrei Atanov et al.

We propose a pre-training strategy called Multi-modal Multi-task Masked Autoencoders (MultiMAE). It differs from standard Masked Autoencoding in two key aspects: I) it can optionally accept additional modalities of information in the input besides the RGB image (hence "multi-modal"), and II) its training objective accordingly includes predicting multiple outputs besides the RGB image (hence "multi-task"). We make use of masking (across image patches and input modalities) to make training MultiMAE tractable as well as to ensure cross-modality predictive coding is indeed learned by the network. We show this pre-training strategy leads to a flexible, simple, and efficient framework with improved transfer results to downstream tasks. In particular, the same exact pre-trained network can be flexibly used when additional information besides RGB images is available or when no information other than RGB is available - in all configurations yielding competitive to or significantly better results than the baselines. To avoid needing training datasets with multiple modalities and tasks, we train MultiMAE entirely using pseudo labeling, which makes the framework widely applicable to any RGB dataset. The experiments are performed on multiple transfer tasks (image classification, semantic segmentation, depth estimation) and datasets (ImageNet, ADE20K, Taskonomy, Hypersim, NYUv2). The results show an intriguingly impressive capability by the model in cross-modal/task predictive coding and transfer.

Oğuzhan Fatih Kar, Teresa Yeo, Andrei Atanov et al.

We introduce a set of image transformations that can be used as corruptions to evaluate the robustness of models as well as data augmentation mechanisms for training neural networks. The primary distinction of the proposed transformations is that, unlike existing approaches such as Common Corruptions, the geometry of the scene is incorporated in the transformations -- thus leading to corruptions that are more likely to occur in the real world. We also introduce a set of semantic corruptions (e.g. natural object occlusions). We show these transformations are `efficient' (can be computed on-the-fly), `extendable' (can be applied on most image datasets), expose vulnerability of existing models, and can effectively make models more robust when employed as `3D data augmentation' mechanisms. The evaluations on several tasks and datasets suggest incorporating 3D information into benchmarking and training opens up a promising direction for robustness research.

Marius Memmel, Roman Bachmann, Amir Zamir

Effectively localizing an agent in a realistic, noisy setting is crucial for many embodied vision tasks. Visual Odometry (VO) is a practical substitute for unreliable GPS and compass sensors, especially in indoor environments. While SLAM-based methods show a solid performance without large data requirements, they are less flexible and robust w.r.t. to noise and changes in the sensor suite compared to learning-based approaches. Recent deep VO models, however, limit themselves to a fixed set of input modalities, e.g., RGB and depth, while training on millions of samples. When sensors fail, sensor suites change, or modalities are intentionally looped out due to available resources, e.g., power consumption, the models fail catastrophically. Furthermore, training these models from scratch is even more expensive without simulator access or suitable existing models that can be fine-tuned. While such scenarios get mostly ignored in simulation, they commonly hinder a model's reusability in real-world applications. We propose a Transformer-based modality-invariant VO approach that can deal with diverse or changing sensor suites of navigation agents. Our model outperforms previous methods while training on only a fraction of the data. We hope this method opens the door to a broader range of real-world applications that can benefit from flexible and learned VO models.

Andrei Atanov, Andrei Filatov, Teresa Yeo et al.

When developing deep learning models, we usually decide what task we want to solve then search for a model that generalizes well on the task. An intriguing question would be: what if, instead of fixing the task and searching in the model space, we fix the model and search in the task space? Can we find tasks that the model generalizes on? How do they look, or do they indicate anything? These are the questions we address in this paper. We propose a task discovery framework that automatically finds examples of such tasks via optimizing a generalization-based quantity called agreement score. We demonstrate that one set of images can give rise to many tasks on which neural networks generalize well. These tasks are a reflection of the inductive biases of the learning framework and the statistical patterns present in the data, thus they can make a useful tool for analysing the neural networks and their biases. As an example, we show that the discovered tasks can be used to automatically create adversarial train-test splits which make a model fail at test time, without changing the pixels or labels, but by only selecting how the datapoints should be split between the train and test sets. We end with a discussion on human-interpretability of the discovered tasks.

Teresa Yeo, Oğuzhan Fatih Kar, Zahra Sodagar et al.

We propose a method for adapting neural networks to distribution shifts at test-time. In contrast to training-time robustness mechanisms that attempt to anticipate and counter the shift, we create a closed-loop system and make use of a test-time feedback signal to adapt a network on the fly. We show that this loop can be effectively implemented using a learning-based function, which realizes an amortized optimizer for the network. This leads to an adaptation method, named Rapid Network Adaptation (RNA), that is notably more flexible and orders of magnitude faster than the baselines. Through a broad set of experiments using various adaptation signals and target tasks, we study the efficiency and flexibility of this method. We perform the evaluations using various datasets (Taskonomy, Replica, ScanNet, Hypersim, COCO, ImageNet), tasks (depth, optical flow, semantic segmentation, classification), and distribution shifts (Cross-datasets, 2D and 3D Common Corruptions) with promising results. We end with a discussion on general formulations for handling distribution shifts and our observations from comparing with similar approaches from other domains.

Onur Beker, Mohammad Mohammadi, Amir Zamir

To achieve autonomy in a priori unknown real-world scenarios, agents should be able to: i) act from high-dimensional sensory observations (e.g., images), ii) learn from past experience to adapt and improve, and iii) be capable of long horizon planning. Classical planning algorithms (e.g. PRM, RRT) are proficient at handling long-horizon planning. Deep learning based methods in turn can provide the necessary representations to address the others, by modeling statistical contingencies between observations. In this direction, we introduce a general-purpose planning algorithm called PALMER that combines classical sampling-based planning algorithms with learning-based perceptual representations. For training these perceptual representations, we combine Q-learning with contrastive representation learning to create a latent space where the distance between the embeddings of two states captures how easily an optimal policy can traverse between them. For planning with these perceptual representations, we re-purpose classical sampling-based planning algorithms to retrieve previously observed trajectory segments from a replay buffer and restitch them into approximately optimal paths that connect any given pair of start and goal states. This creates a tight feedback loop between representation learning, memory, reinforcement learning, and sampling-based planning. The end result is an experiential framework for long-horizon planning that is significantly more robust and sample efficient compared to existing methods.

David Mizrahi, Roman Bachmann, Oğuzhan Fatih Kar et al.

Current machine learning models for vision are often highly specialized and limited to a single modality and task. In contrast, recent large language models exhibit a wide range of capabilities, hinting at a possibility for similarly versatile models in computer vision. In this paper, we take a step in this direction and propose a multimodal training scheme called 4M. It consists of training a single unified Transformer encoder-decoder using a masked modeling objective across a wide range of input/output modalities - including text, images, geometric, and semantic modalities, as well as neural network feature maps. 4M achieves scalability by unifying the representation space of all modalities through mapping them into discrete tokens and performing multimodal masked modeling on a small randomized subset of tokens. 4M leads to models that exhibit several key capabilities: (1) they can perform a diverse set of vision tasks out of the box, (2) they excel when fine-tuned for unseen downstream tasks or new input modalities, and (3) they can function as a generative model that can be conditioned on arbitrary modalities, enabling a wide variety of expressive multimodal editing capabilities with remarkable flexibility. Through experimental analyses, we demonstrate the potential of 4M for training versatile and scalable foundation models for vision tasks, setting the stage for further exploration in multimodal learning for vision and other domains.

Roman Bachmann, Oğuzhan Fatih Kar, David Mizrahi et al.

Current multimodal and multitask foundation models like 4M or UnifiedIO show promising results, but in practice their out-of-the-box abilities to accept diverse inputs and perform diverse tasks are limited by the (usually rather small) number of modalities and tasks they are trained on. In this paper, we expand upon the capabilities of them by training a single model on tens of highly diverse modalities and by performing co-training on large-scale multimodal datasets and text corpora. This includes training on several semantic and geometric modalities, feature maps from recent state of the art models like DINOv2 and ImageBind, pseudo labels of specialist models like SAM and 4DHumans, and a range of new modalities that allow for novel ways to interact with the model and steer the generation, for example image metadata or color palettes. A crucial step in this process is performing discrete tokenization on various modalities, whether they are image-like, neural network feature maps, vectors, structured data like instance segmentation or human poses, or data that can be represented as text. Through this, we expand on the out-of-the-box capabilities of multimodal models and specifically show the possibility of training one model to solve at least 3x more tasks/modalities than existing ones and doing so without a loss in performance. This enables more fine-grained and controllable multimodal generation capabilities and allows us to study the distillation of models trained on diverse data and objectives into a unified model. We successfully scale the training to a three billion parameter model using tens of modalities and different datasets. The resulting models and training code are open sourced at 4m.epfl.ch.

Oğuzhan Fatih Kar, Alessio Tonioni, Petra Poklukar et al.

Vision-language models (VLMs) are typically composed of a vision encoder, e.g. CLIP, and a language model (LM) that interprets the encoded features to solve downstream tasks. Despite remarkable progress, VLMs are subject to several shortcomings due to the limited capabilities of vision encoders, e.g. "blindness" to certain image features, visual hallucination, etc. To address these issues, we study broadening the visual encoding capabilities of VLMs. We first comprehensively benchmark several vision encoders with different inductive biases for solving VLM tasks. We observe that there is no single encoding configuration that consistently achieves top performance across different tasks, and encoders with different biases can perform surprisingly similarly. Motivated by this, we introduce a method, named BRAVE, that consolidates features from multiple frozen encoders into a more versatile representation that can be directly fed as the input to a frozen LM. BRAVE achieves state-of-the-art performance on a broad range of captioning and VQA benchmarks and significantly reduces the aforementioned issues of VLMs, while requiring a smaller number of trainable parameters than existing methods and having a more compressed representation. Our results highlight the potential of incorporating different visual biases for a more broad and contextualized visual understanding of VLMs.

Roman Bachmann, Jesse Allardice, David Mizrahi et al.

Image tokenization has enabled major advances in autoregressive image generation by providing compressed, discrete representations that are more efficient to process than raw pixels. While traditional approaches use 2D grid tokenization, recent methods like TiTok have shown that 1D tokenization can achieve high generation quality by eliminating grid redundancies. However, these methods typically use a fixed number of tokens and thus cannot adapt to an image's inherent complexity. We introduce FlexTok, a tokenizer that projects 2D images into variable-length, ordered 1D token sequences. For example, a 256x256 image can be resampled into anywhere from 1 to 256 discrete tokens, hierarchically and semantically compressing its information. By training a rectified flow model as the decoder and using nested dropout, FlexTok produces plausible reconstructions regardless of the chosen token sequence length. We evaluate our approach in an autoregressive generation setting using a simple GPT-style Transformer. On ImageNet, this approach achieves an FID<2 across 8 to 128 tokens, outperforming TiTok and matching state-of-the-art methods with far fewer tokens. We further extend the model to support to text-conditioned image generation and examine how FlexTok relates to traditional 2D tokenization. A key finding is that FlexTok enables next-token prediction to describe images in a coarse-to-fine "visual vocabulary", and that the number of tokens to generate depends on the complexity of the generation task.

Rahul Ramachandran, Ali Garjani, Roman Bachmann et al.

Multimodal foundation models, such as GPT-4o, have recently made remarkable progress, but it is not clear where exactly these models stand in terms of understanding vision. In this paper, we benchmark the performance of popular multimodal foundation models (GPT-4o, o4-mini, Gemini 1.5 Pro and Gemini 2.0 Flash, Claude 3.5 Sonnet, Qwen2-VL, Llama 3.2) on standard computer vision tasks (semantic segmentation, object detection, image classification, depth and surface normal prediction) using established datasets (e.g., COCO, ImageNet and its variants, etc). The main challenges to performing this are: 1) most models are trained to output text and cannot natively express versatile domains, such as segments or 3D geometry, and 2) many leading models are proprietary and accessible only at an API level, i.e., there is no weight access to adapt them. We address these challenges by translating standard vision tasks into equivalent text-promptable and API-compatible tasks via prompt chaining to create a standardized benchmarking framework. We observe that 1) the models are not close to the state-of-the-art specialist models at any task. However, 2) they are respectable generalists; this is remarkable as they are presumably trained on primarily image-text-based tasks. 3) They perform semantic tasks notably better than geometric ones. 4) While the prompt-chaining techniques affect performance, better models exhibit less sensitivity to prompt variations. 5) GPT-4o performs the best among non-reasoning models, securing the top position in 4 out of 6 tasks, 6) reasoning models, e.g. o3, show improvements in geometric tasks, and 7) a preliminary analysis of models with native image generation, like the latest GPT-4o, shows they exhibit quirks like hallucinations and spatial misalignments.

Teresa Yeo, Andrei Atanov, Harold Benoit et al.

We present a method to control a text-to-image generative model to produce training data useful for supervised learning. Unlike previous works that employ an open-loop approach and pre-define prompts to generate new data using either a language model or human expertise, we develop an automated closed-loop system which involves two feedback mechanisms. The first mechanism uses feedback from a given supervised model and finds adversarial prompts that result in image generations that maximize the model loss. While these adversarial prompts result in diverse data informed by the model, they are not informed of the target distribution, which can be inefficient. Therefore, we introduce the second feedback mechanism that guides the generation process towards a certain target distribution. We call the method combining these two mechanisms Guided Adversarial Prompts. We perform our evaluations on different tasks, datasets and architectures, with different types of distribution shifts (spuriously correlated data, unseen domains) and demonstrate the efficiency of the proposed feedback mechanisms compared to open-loop approaches.

Harold Benoit, Liangze Jiang, Andrei Atanov et al. · ibm-research

Supervised learning datasets may contain multiple cues that explain the training set equally well, i.e., learning any of them would lead to the correct predictions on the training data. However, many of them can be spurious, i.e., lose their predictive power under a distribution shift and consequently fail to generalize to out-of-distribution (OOD) data. Recently developed "diversification" methods (Lee et al., 2023; Pagliardini et al., 2023) approach this problem by finding multiple diverse hypotheses that rely on different features. This paper aims to study this class of methods and identify the key components contributing to their OOD generalization abilities. We show that (1) diversification methods are highly sensitive to the distribution of the unlabeled data used for diversification and can underperform significantly when away from a method-specific sweet spot. (2) Diversification alone is insufficient for OOD generalization. The choice of the used learning algorithm, e.g., the model's architecture and pretraining, is crucial. In standard experiments (classification on Waterbirds and Office-Home datasets), using the second-best choice leads to an up to 20\% absolute drop in accuracy. (3) The optimal choice of learning algorithm depends on the unlabeled data and vice versa i.e. they are co-dependent. (4) Finally, we show that, in practice, the above pitfalls cannot be alleviated by increasing the number of diverse hypotheses, the major feature of diversification methods. These findings provide a clearer understanding of the critical design factors influencing the OOD generalization abilities of diversification methods. They can guide practitioners in how to use the existing methods best and guide researchers in developing new, better ones.

Artyom Gadetsky, Andrei Atanov, Yulun Jiang et al.

Recent advances in large language and vision-language models have enabled zero-shot inference, allowing models to solve new tasks without task-specific training. Various adaptation techniques such as prompt engineering, In-Context Learning (ICL), and supervised fine-tuning can further enhance the model's performance on a downstream task, but they require substantial manual effort to construct effective prompts or labeled examples. In this work, we introduce a joint inference framework for fully unsupervised adaptation, eliminating the need for manual prompt engineering and labeled examples. Unlike zero-shot inference, which makes independent predictions, the joint inference makes predictions simultaneously for all inputs in a given task. Since direct joint inference involves computationally expensive optimization, we develop efficient approximation techniques, leading to two unsupervised adaptation methods: unsupervised fine-tuning and unsupervised ICL. We demonstrate the effectiveness of our methods across diverse tasks and models, including language-only Llama-3.1 on natural language processing tasks, reasoning-oriented Qwen2.5-Math on grade school math problems, vision-language OpenFlamingo on vision tasks, and the API-only access GPT-4o model on massive multi-discipline tasks. Our experiments demonstrate substantial improvements over the standard zero-shot approach, including 39% absolute improvement on the challenging GSM8K math reasoning dataset. Remarkably, despite being fully unsupervised, our framework often performs on par with supervised approaches that rely on ground truth labels.

Andrei Atanov, Jiawei Fu, Rishubh Singh et al.

A de facto standard in solving computer vision problems is to use a common high-resolution camera and choose its placement on an agent (i.e., position and orientation) based on human intuition. On the other hand, extremely simple and well-designed visual sensors found throughout nature allow many organisms to perform diverse, complex behaviors. In this work, motivated by these examples, we raise the following questions: 1. How effective simple visual sensors are in solving vision tasks? 2. What role does their design play in their effectiveness? We explore simple sensors with resolutions as low as one-by-one pixel, representing a single photoreceptor First, we demonstrate that just a few photoreceptors can be enough to solve many tasks, such as visual navigation and continuous control, reasonably well, with performance comparable to that of a high-resolution camera. Second, we show that the design of these simple visual sensors plays a crucial role in their ability to provide useful information and successfully solve these tasks. To find a well-performing design, we present a computational design optimization algorithm and evaluate its effectiveness across different tasks and domains, showing promising results. Finally, we perform a human survey to evaluate the effectiveness of intuitive designs devised manually by humans, showing that the computationally found design is among the best designs in most cases.

Yael Vinker, Ehsan Pajouheshgar, Jessica Y. Bo et al.

Abstraction is at the heart of sketching due to the simple and minimal nature of line drawings. Abstraction entails identifying the essential visual properties of an object or scene, which requires semantic understanding and prior knowledge of high-level concepts. Abstract depictions are therefore challenging for artists, and even more so for machines. We present CLIPasso, an object sketching method that can achieve different levels of abstraction, guided by geometric and semantic simplifications. While sketch generation methods often rely on explicit sketch datasets for training, we utilize the remarkable ability of CLIP (Contrastive-Language-Image-Pretraining) to distill semantic concepts from sketches and images alike. We define a sketch as a set of Bézier curves and use a differentiable rasterizer to optimize the parameters of the curves directly with respect to a CLIP-based perceptual loss. The abstraction degree is controlled by varying the number of strokes. The generated sketches demonstrate multiple levels of abstraction while maintaining recognizability, underlying structure, and essential visual components of the subject drawn.

Andrei Atanov, Shijian Xu, Onur Beker et al.

Transfer learning has witnessed remarkable progress in recent years, for example, with the introduction of augmentation-based contrastive self-supervised learning methods. While a number of large-scale empirical studies on the transfer performance of such models have been conducted, there is not yet an agreed-upon set of control baselines, evaluation practices, and metrics to report, which often hinders a nuanced and calibrated understanding of the real efficacy of the methods. We share an evaluation standard that aims to quantify and communicate transfer learning performance in an informative and accessible setup. This is done by baking a number of simple yet critical control baselines in the evaluation method, particularly the blind-guess (quantifying the dataset bias), scratch-model (quantifying the architectural contribution), and maximal-supervision (quantifying the upper-bound). To demonstrate how the evaluation standard can be employed, we provide an example empirical study investigating a few basic questions about self-supervised learning. For example, using this standard, the study shows the effectiveness of existing self-supervised pre-training methods is skewed towards image classification tasks versus dense pixel-wise predictions. In general, we encourage using/reporting the suggested control baselines in evaluating transfer learning in order to gain a more meaningful and informative understanding.

Ainaz Eftekhar, Alexander Sax, Roman Bachmann et al.

This paper introduces a pipeline to parametrically sample and render multi-task vision datasets from comprehensive 3D scans from the real world. Changing the sampling parameters allows one to "steer" the generated datasets to emphasize specific information. In addition to enabling interesting lines of research, we show the tooling and generated data suffice to train robust vision models. Common architectures trained on a generated starter dataset reached state-of-the-art performance on multiple common vision tasks and benchmarks, despite having seen no benchmark or non-pipeline data. The depth estimation network outperforms MiDaS and the surface normal estimation network is the first to achieve human-level performance for in-the-wild surface normal estimation -- at least according to one metric on the OASIS benchmark. The Dockerized pipeline with CLI, the (mostly python) code, PyTorch dataloaders for the generated data, the generated starter dataset, download scripts and other utilities are available through our project website, https://omnidata.vision.

Teresa Yeo, Oğuzhan Fatih Kar, Alexander Sax et al.

We present a method for making neural network predictions robust to shifts from the training data distribution. The proposed method is based on making predictions via a diverse set of cues (called 'middle domains') and ensembling them into one strong prediction. The premise of the idea is that predictions made via different cues respond differently to a distribution shift, hence one should be able to merge them into one robust final prediction. We perform the merging in a straightforward but principled manner based on the uncertainty associated with each prediction. The evaluations are performed using multiple tasks and datasets (Taskonomy, Replica, ImageNet, CIFAR) under a wide range of adversarial and non-adversarial distribution shifts which demonstrate the proposed method is considerably more robust than its standard learning counterpart, conventional deep ensembles, and several other baselines.

Bryan Chen, Alexander Sax, Gene Lewis et al.

Vision-based robotics often separates the control loop into one module for perception and a separate module for control. It is possible to train the whole system end-to-end (e.g. with deep RL), but doing it "from scratch" comes with a high sample complexity cost and the final result is often brittle, failing unexpectedly if the test environment differs from that of training. We study the effects of using mid-level visual representations (features learned asynchronously for traditional computer vision objectives), as a generic and easy-to-decode perceptual state in an end-to-end RL framework. Mid-level representations encode invariances about the world, and we show that they aid generalization, improve sample complexity, and lead to a higher final performance. Compared to other approaches for incorporating invariances, such as domain randomization, asynchronously trained mid-level representations scale better: both to harder problems and to larger domain shifts. In practice, this means that mid-level representations could be used to successfully train policies for tasks where domain randomization and learning-from-scratch failed. We report results on both manipulation and navigation tasks, and for navigation include zero-shot sim-to-real experiments on real robots.

Amir Zamir, Alexander Sax, Teresa Yeo et al.

Visual perception entails solving a wide set of tasks, e.g., object detection, depth estimation, etc. The predictions made for multiple tasks from the same image are not independent, and therefore, are expected to be consistent. We propose a broadly applicable and fully computational method for augmenting learning with Cross-Task Consistency. The proposed formulation is based on inference-path invariance over a graph of arbitrary tasks. We observe that learning with cross-task consistency leads to more accurate predictions and better generalization to out-of-distribution inputs. This framework also leads to an informative unsupervised quantity, called Consistency Energy, based on measuring the intrinsic consistency of the system. Consistency Energy correlates well with the supervised error (r=0.67), thus it can be employed as an unsupervised confidence metric as well as for detection of out-of-distribution inputs (ROC-AUC=0.95). The evaluations are performed on multiple datasets, including Taskonomy, Replica, CocoDoom, and ApolloScape, and they benchmark cross-task consistency versus various baselines including conventional multi-task learning, cycle consistency, and analytical consistency.

Jeffrey O Zhang, Alexander Sax, Amir Zamir et al.

When training a neural network for a desired task, one may prefer to adapt a pre-trained network rather than starting from randomly initialized weights. Adaptation can be useful in cases when training data is scarce, when a single learner needs to perform multiple tasks, or when one wishes to encode priors in the network. The most commonly employed approaches for network adaptation are fine-tuning and using the pre-trained network as a fixed feature extractor, among others. In this paper, we propose a straightforward alternative: side-tuning. Side-tuning adapts a pre-trained network by training a lightweight "side" network that is fused with the (unchanged) pre-trained network via summation. This simple method works as well as or better than existing solutions and it resolves some of the basic issues with fine-tuning, fixed features, and other common approaches. In particular, side-tuning is less prone to overfitting, is asymptotically consistent, and does not suffer from catastrophic forgetting in incremental learning. We demonstrate the performance of side-tuning under a diverse set of scenarios, including incremental learning (iCIFAR, iTaskonomy), reinforcement learning, imitation learning (visual navigation in Habitat), NLP question-answering (SQuAD v2), and single-task transfer learning (Taskonomy), with consistently promising results.

Alexander Sax, Jeffrey O. Zhang, Bradley Emi et al.

How much does having visual priors about the world (e.g. the fact that the world is 3D) assist in learning to perform downstream motor tasks (e.g. navigating a complex environment)? What are the consequences of not utilizing such visual priors in learning? We study these questions by integrating a generic perceptual skill set (a distance estimator, an edge detector, etc.) within a reinforcement learning framework (see Fig. 1). This skill set ("mid-level vision") provides the policy with a more processed state of the world compared to raw images. Our large-scale study demonstrates that using mid-level vision results in policies that learn faster, generalize better, and achieve higher final performance, when compared to learning from scratch and/or using state-of-the-art visual and non-visual representation learning methods. We show that conventional computer vision objectives are particularly effective in this regard and can be conveniently integrated into reinforcement learning frameworks. Finally, we found that no single visual representation was universally useful for all downstream tasks, hence we computationally derive a task-agnostic set of representations optimized to support arbitrary downstream tasks.

Iro Armeni, Zhi-Yang He, JunYoung Gwak et al.

A comprehensive semantic understanding of a scene is important for many applications - but in what space should diverse semantic information (e.g., objects, scene categories, material types, texture, etc.) be grounded and what should be its structure? Aspiring to have one unified structure that hosts diverse types of semantics, we follow the Scene Graph paradigm in 3D, generating a 3D Scene Graph. Given a 3D mesh and registered panoramic images, we construct a graph that spans the entire building and includes semantics on objects (e.g., class, material, and other attributes), rooms (e.g., scene category, volume, etc.) and cameras (e.g., location, etc.), as well as the relationships among these entities. However, this process is prohibitively labor heavy if done manually. To alleviate this we devise a semi-automatic framework that employs existing detection methods and enhances them using two main constraints: I. framing of query images sampled on panoramas to maximize the performance of 2D detectors, and II. multi-view consistency enforcement across 2D detections that originate in different camera locations.

Trevor Standley, Amir R. Zamir, Dawn Chen et al.

Many computer vision applications require solving multiple tasks in real-time. A neural network can be trained to solve multiple tasks simultaneously using multi-task learning. This can save computation at inference time as only a single network needs to be evaluated. Unfortunately, this often leads to inferior overall performance as task objectives can compete, which consequently poses the question: which tasks should and should not be learned together in one network when employing multi-task learning? We study task cooperation and competition in several different learning settings and propose a framework for assigning tasks to a few neural networks such that cooperating tasks are computed by the same neural network, while competing tasks are computed by different networks. Our framework offers a time-accuracy trade-off and can produce better accuracy using less inference time than not only a single large multi-task neural network but also many single-task networks.

Alexander Sax, Bradley Emi, Amir R. Zamir et al.

How much does having visual priors about the world (e.g. the fact that the world is 3D) assist in learning to perform downstream motor tasks (e.g. delivering a package)? We study this question by integrating a generic perceptual skill set (e.g. a distance estimator, an edge detector, etc.) within a reinforcement learning framework--see Figure 1. This skill set (hereafter mid-level perception) provides the policy with a more processed state of the world compared to raw images. We find that using a mid-level perception confers significant advantages over training end-to-end from scratch (i.e. not leveraging priors) in navigation-oriented tasks. Agents are able to generalize to situations where the from-scratch approach fails and training becomes significantly more sample efficient. However, we show that realizing these gains requires careful selection of the mid-level perceptual skills. Therefore, we refine our findings into an efficient max-coverage feature set that can be adopted in lieu of raw images. We perform our study in completely separate buildings for training and testing and compare against visually blind baseline policies and state-of-the-art feature learning methods.

Fei Xia, Amir Zamir, Zhi-Yang He et al.

Developing visual perception models for active agents and sensorimotor control are cumbersome to be done in the physical world, as existing algorithms are too slow to efficiently learn in real-time and robots are fragile and costly. This has given rise to learning-in-simulation which consequently casts a question on whether the results transfer to real-world. In this paper, we are concerned with the problem of developing real-world perception for active agents, propose Gibson Virtual Environment for this purpose, and showcase sample perceptual tasks learned therein. Gibson is based on virtualizing real spaces, rather than using artificially designed ones, and currently includes over 1400 floor spaces from 572 full buildings. The main characteristics of Gibson are: I. being from the real-world and reflecting its semantic complexity, II. having an internal synthesis mechanism, "Goggles", enabling deploying the trained models in real-world without needing further domain adaptation, III. embodiment of agents and making them subject to constraints of physics and space.

Peter Anderson, Angel Chang, Devendra Singh Chaplot et al.

Skillful mobile operation in three-dimensional environments is a primary topic of study in Artificial Intelligence. The past two years have seen a surge of creative work on navigation. This creative output has produced a plethora of sometimes incompatible task definitions and evaluation protocols. To coordinate ongoing and future research in this area, we have convened a working group to study empirical methodology in navigation research. The present document summarizes the consensus recommendations of this working group. We discuss different problem statements and the role of generalization, present evaluation measures, and provide standard scenarios that can be used for benchmarking.

Amir Zamir, Alexander Sax, William Shen et al.

Do visual tasks have a relationship, or are they unrelated? For instance, could having surface normals simplify estimating the depth of an image? Intuition answers these questions positively, implying existence of a structure among visual tasks. Knowing this structure has notable values; it is the concept underlying transfer learning and provides a principled way for identifying redundancies across tasks, e.g., to seamlessly reuse supervision among related tasks or solve many tasks in one system without piling up the complexity. We proposes a fully computational approach for modeling the structure of space of visual tasks. This is done via finding (first and higher-order) transfer learning dependencies across a dictionary of twenty six 2D, 2.5D, 3D, and semantic tasks in a latent space. The product is a computational taxonomic map for task transfer learning. We study the consequences of this structure, e.g. nontrivial emerged relationships, and exploit them to reduce the demand for labeled data. For example, we show that the total number of labeled datapoints needed for solving a set of 10 tasks can be reduced by roughly 2/3 (compared to training independently) while keeping the performance nearly the same. We provide a set of tools for computing and probing this taxonomical structure including a solver that users can employ to devise efficient supervision policies for their use cases.

Amir R. Zamir, Tilman Wekel, Pulkit Argrawal et al.

Though a large body of computer vision research has investigated developing generic semantic representations, efforts towards developing a similar representation for 3D has been limited. In this paper, we learn a generic 3D representation through solving a set of foundational proxy 3D tasks: object-centric camera pose estimation and wide baseline feature matching. Our method is based upon the premise that by providing supervision over a set of carefully selected foundational tasks, generalization to novel tasks and abstraction capabilities can be achieved. We empirically show that the internal representation of a multi-task ConvNet trained to solve the above core problems generalizes to novel 3D tasks (e.g., scene layout estimation, object pose estimation, surface normal estimation) without the need for fine-tuning and shows traits of abstraction abilities (e.g., cross-modality pose estimation). In the context of the core supervised tasks, we demonstrate our representation achieves state-of-the-art wide baseline feature matching results without requiring apriori rectification (unlike SIFT and the majority of learned features). We also show 6DOF camera pose estimation given a pair local image patches. The accuracy of both supervised tasks come comparable to humans. Finally, we contribute a large-scale dataset composed of object-centric street view scenes along with point correspondences and camera pose information, and conclude with a discussion on the learned representation and open research questions.

Iro Armeni, Sasha Sax, Amir R. Zamir et al.

We present a dataset of large-scale indoor spaces that provides a variety of mutually registered modalities from 2D, 2.5D and 3D domains, with instance-level semantic and geometric annotations. The dataset covers over 6,000m2 and contains over 70,000 RGB images, along with the corresponding depths, surface normals, semantic annotations, global XYZ images (all in forms of both regular and 360° equirectangular images) as well as camera information. It also includes registered raw and semantically annotated 3D meshes and point clouds. The dataset enables development of joint and cross-modal learning models and potentially unsupervised approaches utilizing the regularities present in large-scale indoor spaces. The dataset is available here: http://3Dsemantics.stanford.edu/

Amir R. Zamir, Te-Lin Wu, Lin Sun et al.

Currently, the most successful learning models in computer vision are based on learning successive representations followed by a decision layer. This is usually actualized through feedforward multilayer neural networks, e.g. ConvNets, where each layer forms one of such successive representations. However, an alternative that can achieve the same goal is a feedback based approach in which the representation is formed in an iterative manner based on a feedback received from previous iteration's output. We establish that a feedback based approach has several fundamental advantages over feedforward: it enables making early predictions at the query time, its output naturally conforms to a hierarchical structure in the label space (e.g. a taxonomy), and it provides a new basis for Curriculum Learning. We observe that feedback networks develop a considerably different representation compared to feedforward counterparts, in line with the aforementioned advantages. We put forth a general feedback based learning architecture with the endpoint results on par or better than existing feedforward networks with the addition of the above advantages. We also investigate several mechanisms in feedback architectures (e.g. skip connections in time) and design choices (e.g. feedback length). We hope this study offers new perspectives in quest for more natural and practical learning models.

Ozan Sener, Amir Roshan Zamir, Chenxia Wu et al.

Human communication takes many forms, including speech, text and instructional videos. It typically has an underlying structure, with a starting point, ending, and certain objective steps between them. In this paper, we consider instructional videos where there are tens of millions of them on the Internet. We propose a method for parsing a video into such semantic steps in an unsupervised way. Our method is capable of providing a semantic "storyline" of the video composed of its objective steps. We accomplish this using both visual and language cues in a joint generative model. Our method can also provide a textual description for each of the identified semantic steps and video segments. We evaluate our method on a large number of complex YouTube videos and show that our method discovers semantically correct instructions for a variety of tasks.

Haroon Idrees, Amir R. Zamir, Yu-Gang Jiang et al.

Automatically recognizing and localizing wide ranges of human actions has crucial importance for video understanding. Towards this goal, the THUMOS challenge was introduced in 2013 to serve as a benchmark for action recognition. Until then, video action recognition, including THUMOS challenge, had focused primarily on the classification of pre-segmented (i.e., trimmed) videos, which is an artificial task. In THUMOS 2014, we elevated action recognition to a more practical level by introducing temporally untrimmed videos. These also include `background videos' which share similar scenes and backgrounds as action videos, but are devoid of the specific actions. The three editions of the challenge organized in 2013--2015 have made THUMOS a common benchmark for action classification and detection and the annual challenge is widely attended by teams from around the world. In this paper we describe the THUMOS benchmark in detail and give an overview of data collection and annotation procedures. We present the evaluation protocols used to quantify results in the two THUMOS tasks of action classification and temporal detection. We also present results of submissions to the THUMOS 2015 challenge and review the participating approaches. Additionally, we include a comprehensive empirical study evaluating the differences in action recognition between trimmed and untrimmed videos, and how well methods trained on trimmed videos generalize to untrimmed videos. We conclude by proposing several directions and improvements for future THUMOS challenges.

Ashesh Jain, Amir R. Zamir, Silvio Savarese et al.

Deep Recurrent Neural Network architectures, though remarkably capable at modeling sequences, lack an intuitive high-level spatio-temporal structure. That is while many problems in computer vision inherently have an underlying high-level structure and can benefit from it. Spatio-temporal graphs are a popular tool for imposing such high-level intuitions in the formulation of real world problems. In this paper, we propose an approach for combining the power of high-level spatio-temporal graphs and sequence learning success of Recurrent Neural Networks~(RNNs). We develop a scalable method for casting an arbitrary spatio-temporal graph as a rich RNN mixture that is feedforward, fully differentiable, and jointly trainable. The proposed method is generic and principled as it can be used for transforming any spatio-temporal graph through employing a certain set of well defined steps. The evaluations of the proposed approach on a diverse set of problems, ranging from modeling human motion to object interactions, shows improvement over the state-of-the-art with a large margin. We expect this method to empower new approaches to problem formulation through high-level spatio-temporal graphs and Recurrent Neural Networks.

Francesco Castaldo, Amir Zamir, Roland Angst et al.

Matching cross-view images is challenging because the appearance and viewpoints are significantly different. While low-level features based on gradient orientations or filter responses can drastically vary with such changes in viewpoint, semantic information of images however shows an invariant characteristic in this respect. Consequently, semantically labeled regions can be used for performing cross-view matching. In this paper, we therefore explore this idea and propose an automatic method for detecting and representing the semantic information of an RGB image with the goal of performing cross-view matching with a (non-RGB) geographic information system (GIS). A segmented image forms the input to our system with segments assigned to semantic concepts such as traffic signs, lakes, roads, foliage, etc. We design a descriptor to robustly capture both, the presence of semantic concepts and the spatial layout of those segments. Pairwise distances between the descriptors extracted from the GIS map and the query image are then used to generate a shortlist of the most promising locations with similar semantic concepts in a consistent spatial layout. An experimental evaluation with challenging query images and a large urban area shows promising results.

Tian Lan, Yuke Zhu, Amir Roshan Zamir et al.

Realistic videos of human actions exhibit rich spatiotemporal structures at multiple levels of granularity: an action can always be decomposed into multiple finer-grained elements in both space and time. To capture this intuition, we propose to represent videos by a hierarchy of mid-level action elements (MAEs), where each MAE corresponds to an action-related spatiotemporal segment in the video. We introduce an unsupervised method to generate this representation from videos. Our method is capable of distinguishing action-related segments from background segments and representing actions at multiple spatiotemporal resolutions. Given a set of spatiotemporal segments generated from the training data, we introduce a discriminative clustering algorithm that automatically discovers MAEs at multiple levels of granularity. We develop structured models that capture a rich set of spatial, temporal and hierarchical relations among the segments, where the action label and multiple levels of MAE labels are jointly inferred. The proposed model achieves state-of-the-art performance in multiple action recognition benchmarks. Moreover, we demonstrate the effectiveness of our model in real-world applications such as action recognition in large-scale untrimmed videos and action parsing.

Ozan Sener, Amir Zamir, Silvio Savarese et al.

Human communication typically has an underlying structure. This is reflected in the fact that in many user generated videos, a starting point, ending, and certain objective steps between these two can be identified. In this paper, we propose a method for parsing a video into such semantic steps in an unsupervised way. The proposed method is capable of providing a semantic "storyline" of the video composed of its objective steps. We accomplish this using both visual and language cues in a joint generative model. The proposed method can also provide a textual description for each of the identified semantic steps and video segments. We evaluate this method on a large number of complex YouTube videos and show results of unprecedented quality for this intricate and impactful problem.