Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision

Pre-trained representations are becoming crucial for many NLP and perception tasks. While representation learning in NLP has transitioned to training on raw text without human annotations, visual and vision-language representations still rely heavily on curated training datasets that are expensive or require expert knowledge. For vision applications, representations are mostly learned using datasets with explicit class labels such as ImageNet or OpenImages. For vision-language, popular datasets like Conceptual Captions, MSCOCO, or CLIP all involve a non-trivial data collection (and cleaning) process. This costly curation process limits the size of datasets and hence hinders the scaling of trained models. In this paper, we leverage a noisy dataset of over one billion image alt-text pairs, obtained without expensive filtering or post-processing steps in the Conceptual Captions dataset. A simple dual-encoder architecture learns to align visual and language representations of the image and text pairs using a contrastive loss. We show that the scale of our corpus can make up for its noise and leads to state-of-the-art representations even with such a simple learning scheme. Our visual representation achieves strong performance when transferred to classification tasks such as ImageNet and VTAB. The aligned visual and language representations also set new state-of-the-art results on Flickr30K and MSCOCO benchmarks, even when compared with more sophisticated cross-attention models. The representations also enable cross-modality search with complex text and text + image queries.

Unsupervised Domain Adaptation for Spatio-Temporal Action Localization

Spatio-temporal action localization is an important problem in computer vision that involves detecting where and when activities occur, and therefore requires modeling of both spatial and temporal features. This problem is typically formulated in the context of supervised learning, where the learned classifiers operate on the premise that both training and test data are sampled from the same underlying distribution. However, this assumption does not hold when there is a significant domain shift, leading to poor generalization performance on the test data. To address this, we focus on the hard and novel task of generalizing training models to test samples without access to any labels from the latter for spatio-temporal action localization by proposing an end-to-end unsupervised domain adaptation algorithm. We extend the state-of-the-art object detection framework to localize and classify actions. In order to minimize the domain shift, three domain adaptation modules at image level (temporal and spatial) and instance level (temporal) are designed and integrated. We design a new experimental setup and evaluate the proposed method and different adaptation modules on the UCF-Sports, UCF-101 and JHMDB benchmark datasets. We show that significant performance gain can be achieved when spatial and temporal features are adapted separately, or jointly for the most effective results.

Uncertainty-aware Self-supervised 3D Data Association

3D object trackers usually require training on large amounts of annotated data that is expensive and time-consuming to collect. Instead, we propose leveraging vast unlabeled datasets by self-supervised metric learning of 3D object trackers, with a focus on data association. Large scale annotations for unlabeled data are cheaply obtained by automatic object detection and association across frames. We show how these self-supervised annotations can be used in a principled manner to learn point-cloud embeddings that are effective for 3D tracking. We estimate and incorporate uncertainty in self-supervised tracking to learn more robust embeddings, without needing any labeled data. We design embeddings to differentiate objects across frames, and learn them using uncertainty-aware self-supervised training. Finally, we demonstrate their ability to perform accurate data association across frames, towards effective and accurate 3D tracking. Project videos and code are at https://jianrenw.github.io/Self-Supervised-3D-Data-Association.

Who Make Drivers Stop? Towards Driver-centric Risk Assessment: Risk Object Identification via Causal Inference

We propose a framework based on causal inference for risk object identification, an essential task towards driver-centric risk assessment. In this work, risk objects are defined as objects influencing driver's goal-oriented behavior. There are two limitations of the existing approaches. First, they require strong supervisions such as risk object location or human gaze location. Second, there is no explicit reasoning stage for identifying risk object. To address these issues, the task of identifying causes of driver behavioral change is formalized in the language of functional causal models and interventions. Specifically, we iteratively simulate causal effect by removing an object using the proposed driving model. The risk object is determined as the one causing the most substantial causal effect. We evaluate the proposed framework on the Honda Research Institute Driving Dataset (HDD). The dataset provides the annotation for risk object localization to enable systematic benchmarking with existing approaches. Our framework demonstrates a substantial average performance boost over a strong baseline by 7.5%.

Grounding Human-to-Vehicle Advice for Self-driving Vehicles

Recent success suggests that deep neural control networks are likely to be a key component of self-driving vehicles. These networks are trained on large datasets to imitate human actions, but they lack semantic understanding of image contents. This makes them brittle and potentially unsafe in situations that do not match training data. Here, we propose to address this issue by augmenting training data with natural language advice from a human. Advice includes guidance about what to do and where to attend. We present the first step toward advice giving, where we train an end-to-end vehicle controller that accepts advice. The controller adapts the way it attends to the scene (visual attention) and the control (steering and speed). Attention mechanisms tie controller behavior to salient objects in the advice. We evaluate our model on a novel advisable driving dataset with manually annotated human-to-vehicle advice called Honda Research Institute-Advice Dataset (HAD). We show that taking advice improves the performance of the end-to-end network, while the network cues on a variety of visual features that are provided by advice. The dataset is available at https://usa.honda-ri.com/HAD.

Learning 3D-aware Egocentric Spatial-Temporal Interaction via Graph Convolutional Networks

To enable intelligent automated driving systems, a promising strategy is to understand how human drives and interacts with road users in complicated driving situations. In this paper, we propose a 3D-aware egocentric spatial-temporal interaction framework for automated driving applications. Graph convolution networks (GCN) is devised for interaction modeling. We introduce three novel concepts into GCN. First, we decompose egocentric interactions into ego-thing and ego-stuff interaction, modeled by two GCNs. In both GCNs, ego nodes are introduced to encode the interaction between thing objects (e.g., car and pedestrian), and interaction between stuff objects (e.g., lane marking and traffic light). Second, objects' 3D locations are explicitly incorporated into GCN to better model egocentric interactions. Third, to implement ego-stuff interaction in GCN, we propose a MaskAlign operation to extract features for irregular objects. We validate the proposed framework on tactical driver behavior recognition. Extensive experiments are conducted using Honda Research Institute Driving Dataset, the largest dataset with diverse tactical driver behavior annotations. Our framework demonstrates substantial performance boost over baselines on the two experimental settings by 3.9% and 6.0%, respectively. Furthermore, we visualize the learned affinity matrices, which encode ego-thing and ego-stuff interactions, to showcase the proposed framework can capture interactions effectively.

The H3D Dataset for Full-Surround 3D Multi-Object Detection and Tracking in Crowded Urban Scenes

3D multi-object detection and tracking are crucial for traffic scene understanding. However, the community pays less attention to these areas due to the lack of a standardized benchmark dataset to advance the field. Moreover, existing datasets (e.g., KITTI) do not provide sufficient data and labels to tackle challenging scenes where highly interactive and occluded traffic participants are present. To address the issues, we present the Honda Research Institute 3D Dataset (H3D), a large-scale full-surround 3D multi-object detection and tracking dataset collected using a 3D LiDAR scanner. H3D comprises of 160 crowded and highly interactive traffic scenes with a total of 1 million labeled instances in 27,721 frames. With unique dataset size, rich annotations, and complex scenes, H3D is gathered to stimulate research on full-surround 3D multi-object detection and tracking. To effectively and efficiently annotate a large-scale 3D point cloud dataset, we propose a labeling methodology to speed up the overall annotation cycle. A standardized benchmark is created to evaluate full-surround 3D multi-object detection and tracking algorithms. 3D object detection and tracking algorithms are trained and tested on H3D. Finally, sources of errors are discussed for the development of future algorithms.

Graph-RISE: Graph-Regularized Image Semantic Embedding

Learning image representations to capture fine-grained semantics has been a challenging and important task enabling many applications such as image search and clustering. In this paper, we present Graph-Regularized Image Semantic Embedding (Graph-RISE), a large-scale neural graph learning framework that allows us to train embeddings to discriminate an unprecedented O(40M) ultra-fine-grained semantic labels. Graph-RISE outperforms state-of-the-art image embedding algorithms on several evaluation tasks, including image classification and triplet ranking. We provide case studies to demonstrate that, qualitatively, image retrieval based on Graph-RISE effectively captures semantics and, compared to the state-of-the-art, differentiates nuances at levels that are closer to human-perception.

Unsupervised Data Uncertainty Learning in Visual Retrieval Systems

We introduce an unsupervised formulation to estimate heteroscedastic uncertainty in retrieval systems. We propose an extension to triplet loss that models data uncertainty for each input. Besides improving performance, our formulation models local noise in the embedding space. It quantifies input uncertainty and thus enhances interpretability of the system. This helps identify noisy observations in query and search databases. Evaluation on both image and video retrieval applications highlight the utility of our approach. We highlight our efficiency in modeling local noise using two real-world datasets: Clothing1M and Honda Driving datasets. Qualitative results illustrate our ability in identifying confusing scenarios in various domains. Uncertainty learning also enables data cleaning by detecting noisy training labels.

In Defense of the Triplet Loss for Visual Recognition

We employ triplet loss as a space embedding regularizer to boost classification performance. Standard architectures, like ResNet and DesneNet, are extended to support both losses with minimal hyper-parameter tuning. This promotes generality while fine-tuning pretrained networks. Triplet loss is a powerful surrogate for recently proposed embedding regularizers. Yet, it is avoided for large batch-size requirement and high computational cost. Through our experiments, we re-assess these assumptions. During inference, our network supports both classification and embedding tasks without any computational overhead. Quantitative evaluation highlights how our approach compares favorably to the existing state of the art on multiple fine-grained recognition datasets. Further evaluation on an imbalanced video dataset achieves significant improvement (>7%). Beyond boosting efficiency, triplet loss brings retrieval and interpretability to classification models.