Instructional videos are an important resource to learn procedural tasks from human demonstrations. However, the instruction steps in such videos are typically short and sparse, with most of the video being irrelevant to the procedure. This motivates the need to temporally localize the instruction steps in such videos, i.e. the task called key-step localization. Traditional methods for key-step localization require video-level human annotations and thus do not scale to large datasets. In this work, we tackle the problem with no human supervision and introduce StepFormer, a self-supervised model that discovers and localizes instruction steps in a video. StepFormer is a transformer decoder that attends to the video with learnable queries, and produces a sequence of slots capturing the key-steps in the video. We train our system on a large dataset of instructional videos, using their automatically-generated subtitles as the only source of supervision. In particular, we supervise our system with a sequence of text narrations using an order-aware loss function that filters out irrelevant phrases. We show that our model outperforms all previous unsupervised and weakly-supervised approaches on step detection and localization by a large margin on three challenging benchmarks. Moreover, our model demonstrates an emergent property to solve zero-shot multi-step localization and outperforms all relevant baselines at this task.

We study the problem of future step anticipation in procedural videos. Given a video of an ongoing procedural activity, we predict a plausible next procedure step described in rich natural language. While most previous work focuses on the problem of data scarcity in procedural video datasets, another core challenge of future anticipation is how to account for multiple plausible future realizations in natural settings. This problem has been largely overlooked in previous work. To address this challenge, we frame future step prediction as modelling the distribution of all possible candidates for the next step. Specifically, we design a generative model that takes a series of video clips as input, and generates multiple plausible and diverse candidates (in natural language) for the next step. Following previous work, we side-step the video annotation scarcity by pretraining our model on a large text-based corpus of procedural activities, and then transfer the model to the video domain. Our experiments, both in textual and video domains, show that our model captures diversity in the next step prediction and generates multiple plausible future predictions. Moreover, our model establishes new state-of-the-art results on YouCookII, where it outperforms existing baselines on the next step anticipation. Finally, we also show that our model can successfully transfer from text to the video domain zero-shot, i.e., without fine-tuning or adaptation, and produces good-quality future step predictions from video.

When humans perform a task with an articulated object, they interact with the object only in a handful of ways, while the space of all possible interactions is nearly endless. This is because humans have prior knowledge about what interactions are likely to be successful, i.e., to open a new door we first try the handle. While learning such priors without supervision is easy for humans, it is notoriously hard for machines. In this work, we tackle unsupervised learning of priors of useful interactions with articulated objects, which we call interaction modes. In contrast to the prior art, we use no supervision or privileged information; we only assume access to the depth sensor in the simulator to learn the interaction modes. More precisely, we define a successful interaction as the one changing the visual environment substantially and learn a generative model of such interactions, that can be conditioned on the desired goal state of the object. In our experiments, we show that our model covers most of the human interaction modes, outperforms existing state-of-the-art methods for affordance learning, and can generalize to objects never seen during training. Additionally, we show promising results in the goal-conditional setup, where our model can be quickly fine-tuned to perform a given task. We show in the experiments that such affordance learning predicts interaction which covers most modes of interaction for the querying articulated object and can be fine-tuned to a goal-conditional model. For supplementary: https://actaim.github.io/.

Analyzing reliability of Machine Vision Components (MVC) against scene changes (such as rain or fog) in their operational environment is crucial for safety-critical applications. Safety analysis relies on the availability of precisely specified and, ideally, machine-verifiable requirements. The state-of-the-art reliability framework ICRAF developed machine-verifiable requirements obtained using human performance data. However, ICRAF is limited to analyzing reliability of MVCs solving simple vision tasks, such as image classification. Yet, many real-world safety-critical systems require solving more complex vision tasks, such as object detection and instance segmentation. Fortunately, many complex vision tasks (which we call “c-tasks”) can be represented as a sequence of simple vision subtasks. For instance, object detection can be decomposed as object localization followed by classification. Based on this fact, in this paper, we show that the analysis of c-tasks can also be decomposed as a sequential analysis of their simple subtasks, which allows us to apply existing techniques for analyzing simple vision tasks. Specifically, we propose a modular reliability framework, DecompoVision, that decomposes: (1) the problem of solving a c-task, (2) the reliability requirements, and (3) the reliability analysis, and, as a result, provides deeper insights into MVC reliability. DecompoVision extends ICRAF to handle complex vision tasks and enables reuse of existing artifacts across different c-tasks. We capture new reliability gaps by checking our requirements on 13 widely used object detection MVCs, and, for the first time, benchmark segmentation MVCs.

In this paper, we study the problem of procedure planning in instructional videos. Here, an agent must produce a plausible sequence of actions that can transform the environment from a given start to a desired goal state. When learning procedure planning from instructional videos, most recent work leverages intermediate visual observations as supervision, which requires expensive annotation efforts to localize precisely all the instructional steps in training videos. In contrast, we remove the need for expensive temporal video annotations and propose a weakly supervised approach by learning from natural language instructions. Our model is based on a transformer equipped with a memory module, which maps the start and goal observations to a sequence of plausible actions. Furthermore, we augment our model with a probabilistic generative module to capture the uncertainty inherent to procedure planning, an aspect largely overlooked by previous work. We evaluate our model on three datasets and show our weakly-supervised approach outperforms previous fully supervised state-of-the-art models on multiple metrics.

In this work, we consider the problem of sequence-to-sequence alignment for signals containing outliers. Assuming the absence of outliers, the standard Dynamic Time Warping (DTW) algorithm efficiently computes the optimal alignment between two (generally) variable-length sequences. While DTW is robust to temporal shifts and dilations of the signal, it fails to align sequences in a meaningful way in the presence of outliers that can be arbitrarily interspersed in the sequences. To address this problem, we introduce Drop-DTW, a novel algorithm that aligns the common signal between the sequences while automatically dropping the outlier elements from the matching. The entire procedure is implemented as a single dynamic program that is efficient and fully differentiable. In our experiments, we show that Drop-DTW is a robust similarity measure for sequence retrieval and demonstrate its effectiveness as a training loss on diverse applications. With Drop-DTW, we address temporal step localization on instructional videos, representation learning from noisy videos, and cross-modal representation learning for audio-visual retrieval and localization. In all applications, we take a weakly- or unsupervised approach and demonstrate state-of-the-art results under these settings.

Performing data augmentation for learning deep neural networks is known to be important for training visual recognition systems. By artificially increasing the number of training examples, it helps reducing overfitting and improves generalization. While simple image transformations can already improve predictive performance in most vision tasks, larger gains can be obtained by leveraging task-specific prior knowledge. In this work, we consider object detection, semantic and instance segmentation and augment the training images by blending objects in existing scenes, using instance segmentation annotations. We observe that randomly pasting objects on images hurts the performance, unless the object is placed in the right context. To resolve this issue, we propose an explicit context model by using a convolutional neural network, which predicts whether an image region is suitable for placing a given object or not. In our experiments, we show that our approach is able to improve object detection, semantic and instance segmentation on the PASCAL VOC12 and COCO datasets, with significant gains in a limited annotation scenario, i.e., when only one category is annotated. We also show that the method is not limited to datasets that come with expensive pixel-wise instance annotations and can be used when only bounding boxes are available, by employing weakly-supervised learning for instance masks approximation.

Few-shot classification consists of learning a predictive model that is able to effectively adapt to a new class, given only a few annotated samples. To solve this challenging problem, meta-learning has become a popular paradigm that advocates the ability to "learn to adapt''. Recent works have shown, however, that simple learning strategies without meta-learning could be competitive. In this paper, we go a step further and show that by addressing the fundamental high-variance issue of few-shot learning classifiers, it is possible to significantly outperform current meta-learning techniques. Our approach consists of designing an ensemble of deep networks to leverage the variance of the classifiers, and introducing new strategies to encourage the networks to cooperate, while encouraging prediction diversity. Evaluation is conducted on the mini-ImageNet, tiered-ImageNet and CUB datasets, where we show that even a single network obtained by distillation yields state-of-the-art results.