This work addresses the unsupervised adaptation of an existing object detector to a new target domain. We assume that a large number of unlabeled videos from this domain are readily available. We automatically obtain labels on the target data by using high-confidence detections from the existing detector, augmented with hard (misclassified) examples acquired by exploiting temporal cues using a tracker. These automatically-obtained labels are then used for re-training the original model. A modified knowledge distillation loss is proposed, and we investigate several ways of assigning soft-labels to the training examples from the target domain. Our approach is empirically evaluated on challenging face and pedestrian detection tasks: a face detector trained on WIDER-Face, which consists of highquality images crawled from the web, is adapted to a largescale surveillance data set; a pedestrian detector trained on clear, daytime images from the BDD-100K driving data set is adapted to all other scenarios such as rainy, foggy, nighttime. Our results demonstrate the usefulness of incorporating hard examples obtained from tracking, the advantage of using soft-labels via distillation loss versus hard-labels, and show promising performance as a simple method for unsupervised domain adaptation of object detectors, with minimal dependence on hyper-parameters. Code and models are available at

Important gains have recently been obtained in object detection by using training objectives that focus on hard negative examples, i.e., negative examples that are currently rated as positive or ambiguous by the detector. These examples can strongly influence parameters when the network is trained to correct them. Unfortunately, they are often sparse in the training data, and are expensive to obtain. In this work, we show how large numbers of hard negatives can be obtained automatically by analyzing the output of a trained detector on video sequences. In particular, detections that are isolated in time, i.e., that have no associated preceding or following detections, are likely to be hard negatives. We describe simple procedures for mining large numbers of such hard negatives (and also hard positives) from unlabeled video data. Our experiments show that retraining detectors on these automatically obtained examples often significantly improves performance. We present experiments on multiple architectures and multiple data sets, including face detection, pedestrian detection and other object categories.

Figure 1: Clustering results from Hannah and Her Sisters. Each unique color shows a particular cluster. It can be seen that most individuals appear with a consistent color, indicating successful clustering. AbstractWe present an end-to-end system for detecting and clustering faces by identity in full-length movies. Unlike works that start with a predefined set of detected faces, we consider the end-to-end problem of detection and clustering together. We make three separate contributions. First, we combine a state-of-the-art face detector with a generic tracker to extract high quality face tracklets. We then introduce a novel clustering method, motivated by the classic graph theory results of Erdős and Rényi. It is based on the observations that large clusters can be fully connected by joining just a small fraction of their point pairs, while just a single connection between two different people can lead to poor clustering results. This suggests clustering using a verification system with very few false positives but perhaps moderate recall. We introduce a novel verification method, rank-1 counts verification, that has this property, and use it in a link-based clustering scheme. Finally, we define a novel end-to-end detection and clustering evaluation metric allowing us to assess the accuracy of the entire end-to-end system. We present state-of-the-art results on multiple video data sets and also on standard face databases.

No abstract

No abstract

In contrast to recent advances focusing on highlevel representation learning across modalities, in this work we present a self-supervised learning framework that is able to learn a representation that captures finer levels of granularity across different modalities such as concepts or events represented by visual objects or spoken words. Our framework relies on a discretized embedding space created via vector quantization that is shared across different modalities. Beyond the shared embedding space, we propose a Cross-Modal Code Matching objective that forces the representations from different views (modalities) to have a similar distribution over the discrete embedding space such that cross-modal objects/actions localization can be performed without direct supervision. We show that the proposed discretized multi-modal finegrained representation (e.g., pixel/word/frame) can complement high-level summary representations (e.g., video/sentence/waveform) for improved performance on cross-modal retrieval tasks. We also observe that the discretized representation uses individual clusters to represent the same semantic concept across modalities.

No abstract

Grasping the meaning of everyday visual events is a fundamental feat of human intelligence that hinges on diverse neural processes ranging from vision to higher-level cognition. Deciphering the neural basis of visual event understanding requires rich, extensive, and appropriately designed experimental data. However, this type of data is hitherto missing. To fill this gap, we introduce the BOLD Moments Dataset (BMD), a large dataset of whole-brain fMRI responses to over 1,000 short (3s) naturalistic video clips and accompanying metadata. We show visual events interface with an array of processes, extending even to memory, and we reveal a match in hierarchical processing between brains and video-computable deep neural networks. Furthermore, we showcase that BMD successfully captures temporal dynamics of visual events at second resolution. BMD thus establishes a critical groundwork for investigations of the neural basis of visual event understanding.