Author: Ronghang Hu

Ronghang Hu, Amanpreet Singh

Abstract

We propose UniT, a Unified Transformer model to simultaneously learn the most prominent tasks across different domains, ranging from object detection to natural language understanding and multimodal reasoning. Based on the transformer encoder-decoder architecture, our UniT model encodes each input modality with an encoder and makes predictions on each task with a shared decoder over the encoded input representations, followed by task-specific output heads. The entire model is jointly trained end-to-end with losses from each task. Compared to previous efforts on multi-task learning with transformers, we share the same model parameters across all tasks instead of separately fine-tuning task-specific models and handle a much higher variety of tasks across different domains. In our experiments, we learn 7 tasks jointly over 8 datasets, achieving strong performance on each task with 87.5% fewer parameters.

Publications

• R. Hu, A. Singh. UniT: Multimodal Multitask Learning with a Unified Transformer. arXiv preprint arXiv:2102.10772, 2021.
  (PDF)

@article{hu2021unit,
  title={UniT: Multimodal multitask learning with a unified transformer},
  author={Hu, Ronghang and Singh, Amanpreet},
  journal={arXiv preprint arXiv:2102.10772},
  year={2021}
}

Code

• code will be released in MMF

Ronghang Hu, Amanpreet Singh, Trevor Darrell, Marcus Rohrbach

Abstract

Many visual scenes contain text that carries crucial information, and it is thus essential to understand text in images for downstream reasoning tasks. For example, a deep water label on a warning sign warns people about the danger in the scene. Recent work has explored the TextVQA task that requires reading and understanding text in images to answer a question. However, existing approaches for TextVQA are mostly based on custom pairwise fusion mechanisms between a pair of two modalities and are restricted to a single prediction step by casting TextVQA as a classification task. In this work, we propose a novel model for the TextVQA task based on a multimodal transformer architecture accompanied by a rich representation for text in images. Our model naturally fuses different modalities homogeneously by embedding them into a common semantic space where self-attention is applied to model inter- and intra- modality context. Furthermore, it enables iterative answer decoding with a dynamic pointer network, allowing the model to form an answer through multi-step prediction instead of one-step classification. Our model outperforms existing approaches on three benchmark datasets for the TextVQA task by a large margin.

Publications

• R. Hu, A. Singh, T. Darrell, M. Rohrbach, Iterative Answer Prediction with Pointer-Augmented Multimodal Transformers for TextVQA. in CVPR, 2020
  (PDF)

@inproceedings{hu2020iterative,
  title={Iterative Answer Prediction with Pointer-Augmented Multimodal Transformers for TextVQA},
  author={Hu, Ronghang and Singh, Amanpreet and Darrell, Trevor and Rohrbach, Marcus},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  year={2020}
}

Code

• Latest code available in MMF
• Older version (in PyTorch, built upon Pythia) available at here

Ronghang Hu, Anna Rohrbach, Trevor Darrell, Kate Saenko

Abstract

Solving grounded language tasks often requires reasoning about relationships between objects in the context of a given task. For example, to answer the question "What color is the mug on the plate?" we must check the color of the specific mug that satisfies the "on" relationship with respect to the plate. Recent work has proposed various methods capable of complex relational reasoning. However, most of their power is in the inference structure, while the scene is represented with simple local appearance features. In this paper, we take an alternate approach and build contextualized representations for objects in a visual scene to support relational reasoning. We propose a general framework of Language-Conditioned Graph Networks (LCGN), where each node represents an object, and is described by a context-aware representation from related objects through iterative message passing conditioned on the textual input. E.g., conditioning on the "on" relationship to the plate, the object "mug" gathers messages from the object "plate" to update its representation to "mug on the plate", which can be easily consumed by a simple classifier for answer prediction. We experimentally show that our LCGN approach effectively supports relational reasoning and improves performance across several tasks and datasets.

Publications

• R. Hu, A. Rohrbach, T. Darrell, K. Saenko, Language-Conditioned Graph Networks for Relational Reasoning. in ICCV, 2019
  (PDF)

@inproceedings{hu2019language,
  title={Language-Conditioned Graph Networks for Relational Reasoning},
  author={Hu, Ronghang and Rohrbach, Anna and Darrell, Trevor and Saenko, Kate},
  booktitle={Proceedings of the IEEE International Conference on Computer Vision (ICCV)},
  year={2019}
}

Code

• Code (available in both TensorFlow and PyTorch) available at here.
• We also release our simple but well-performing "single-hop" baseline for the GQA dataset in a standalone repo. This "single-hop" model can serve as a basis for developing more complicated models.

Ronghang Hu, Jacob Andreas, Trevor Darrell, Kate Saenko

Abstract

In complex inferential tasks like question answering, machine learning models must confront two challenges: the need to implement a compositional reasoning process, and, in many applications, the need for this reasoning process to be interpretable to assist users in both development and prediction. Existing models designed to produce interpretable traces of their decision-making process typically require these traces to be supervised at training time. In this paper, we present a novel neural modular approach that performs compositional reasoning by automatically inducing a desired sub-task decomposition without relying on strong supervision. Our model allows linking different reasoning tasks though shared modules that handle common routines across tasks. Experiments show that the model is more interpretable to human evaluators compared to other state-of-the-art models: users can better understand the model's underlying reasoning procedure and predict when it will succeed or fail based on observing its intermediate outputs.

Publications

• R. Hu, J. Andreas, T. Darrell, K. Saenko, Explainable Neural Computation via Stack Neural Module Networks. in ECCV, 2018
  (PDF)

@inproceedings{hu2018explainable,
  title={Explainable Neural Computation via Stack Neural Module Networks},
  author={Hu, Ronghang and Andreas, Jacob and Darrell, Trevor and Saenko, Kate},
  booktitle={Proceedings of the European Conference on Computer Vision (ECCV)},
  year={2018}
}

Datasets

• The CLEVR-Ref dataset used in this work can be downloaded here (19 GB; MD5 checksum: 947c0b42eba0eebaf6df9cc483ee7a0b).

Code

• Code (in TensorFlow) available at here.

Daniel Fried, Ronghang Hu, Volkan Cirik, Anna Rohrbach, Jacob Andreas, Louis-Philippe Morency, Taylor Berg-Kirkpatrick, Kate Saenko, Dan Klein, Trevor Darrell

Abstract

Navigation guided by natural language instructions presents a challenging reasoning problem for instruction followers. Natural language instructions typically identify only a few high-level decisions and landmarks rather than complete low-level motor behaviors; much of the missing information must be inferred based on perceptual context. In machine learning settings, this presents a double challenge: it is difficult to collect enough annotated data to enable learning of this reasoning process from scratch, and empirically difficult to implement the reasoning process using generic sequence models. Here we describe an approach to vision-and-language navigation that addresses both these issues with an embedded speaker model. We use this speaker model to synthesize new instructions for data augmentation and to implement pragmatic reasoning for evaluating candidate action sequences. Both steps are supported by a panoramic action space that reflects the granularity of human-generated instructions. Experiments show that all three pieces of this approach---speaker-driven data augmentation, pragmatic reasoning and panoramic action space---dramatically improve the performance of a baseline instruction follower, more than doubling the success rate over the best existing approach on a standard benchmark.

Publications

• D. Fried*, R. Hu*, V. Cirik*, A. Rohrbach, J. Andreas, L.-P. Morency, T. Berg-Kirkpatrick, K. Saenko, D. Klein**, T. Darrell**, Speaker-Follower Models for Vision-and-Language Navigation. in NeurIPS, 2018 (*, **: indicates equal contribution)
  (PDF)

@inproceedings{fried2018speaker,
  title={Speaker-Follower Models for Vision-and-Language Navigation},
  author={Fried, Daniel and Hu, Ronghang and Cirik, Volkan and Rohrbach, Anna and Andreas, Jacob and Morency, Louis-Philippe and Berg-Kirkpatrick, Taylor and Saenko, Kate and Klein, Dan and Darrell, Trevor},
  booktitle={Neural Information Processing Systems (NeurIPS)},
  year={2018}
}

Augmented Data on R2R

• If you only want to use our data augmentation on the R2R dataset but don't need our models, you can directly download our augmented data on R2R (JSON file containing synthetic data generated by our speaker model) here. This JSON file is in the same format as the original R2R dataset, with one synthetic instruction per sampled new trajectory. Note that we first trained on the combination of the original and the augmented data, and then fine-tuned on the original training data.

Code

• Code (in PyTorch) available at here.

Ronghang Hu, Piotr Dollár, Kaiming He, Trevor Darrell, Ross Girshick

Abstract

Existing methods for object instance segmentation require all training instances to be labeled with segmentation masks. This requirement makes it expensive to annotate new categories and has restricted instance segmentation models to ~100 well-annotated classes. The goal of this paper is to propose a new partially supervised training paradigm, together with a novel weight transfer function, that enables training instance segmentation models over a large set of categories for which all have box annotations, but only a small fraction have mask annotations. These contributions allow us to train Mask R-CNN to detect and segment 3000 visual concepts using box annotations from the Visual Genome dataset and mask annotations from the 80 classes in the COCO dataset. We carefully evaluate our proposed approach in a controlled study on the COCO dataset. This work is a first step towards instance segmentation models that have broad comprehension of the visual world.

Publications

• R. Hu, P. Dollár, K. He, T. Darrell, R. Girshick, Learning to Segment Every Thing. in CVPR, 2018
  (PDF)

@inproceedings{hu2018learning,
  title={Learning to Segment Every Thing},
  author={Hu, Ronghang and Dollár, Piotr and He, Kaiming and Darrell, Trevor and Girshick, Ross},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  year={2018}
}

Code

• Code (in Caffe2, based on Detectron) available at here.

Ronghang Hu, Jacob Andreas, Marcus Rohrbach, Trevor Darrell, Kate Saenko

Abstract

Natural language questions are inherently compositional, and many are most easily answered by reasoning about their decomposition into modular sub-problems. For example, to answer "is there an equal number of balls and boxes?" we can look for balls, look for boxes, count them, and compare the results. The recently proposed Neural Module Network (NMN) architecture implements this approach to question answering by parsing questions into linguistic substructures and assembling question-specific deep networks from smaller modules that each solve one subtask. However, existing NMN implementations rely on brittle off-the-shelf parsers, and are restricted to the module configurations proposed by these parsers rather than learning them from data. In this paper, we propose End-to-End Module Networks (N2NMNs), which learn to reason by directly predicting instance-specific network layouts without the aid of a parser. Our model learns to generate network structures (by imitating expert demonstrations) while simultaneously learning network parameters (using the downstream task loss). Experimental results on the new CLEVR dataset targeted at compositional question answering show that N2NMNs achieve an error reduction of nearly 50% relative to state-of-the-art attentional approaches, while discovering interpretable network architectures specialized for each question.

Method

Our N2NMN model consists of two main components: a set of co-attentive neural modules that provide parameterized functions for solving sub-tasks, and a layout policy to predict a question-specific layout from which a neural network is dynamically assembled. Check out the following video for a detailed introduction of our model.

Our model breaks down complex reasoning problems posed in textual questions into a few sub-tasks connected together, and learns to predict a suitable layout expression for each question using our layout policy implemented with a sequence-to-sequence RNN. Different from previous work, we use a new module parameterization that uses a soft attention over question words rather than hard-coded word assignments. During training, the model can be first trained with behavior cloning from an expert layout policy, and further optimized end-to-end using reinforcement learning.

Publications

• R. Hu, J. Andreas, M. Rohrbach, T. Darrell, K. Saenko, Learning to Reason: End-to-End Module Networks for Visual Question Answering. in ICCV, 2017
  (PDF)

@inproceedings{hu2017learning,
  title={Learning to Reason: End-to-End Module Networks for Visual Question Answering},
  author={Hu, Ronghang and Andreas, Jacob and Rohrbach, Marcus and Darrell, Trevor and Saenko, Kate},
  booktitle={Proceedings of the IEEE International Conference on Computer Vision (ICCV)},
  year={2017}
}

Code

• Code (in TensorFlow) available at here.

Ronghang Hu, Marcus Rohrbach, Jacob Andreas, Trevor Darrell, Kate Saenko

Abstract

People often refer to entities in an image in terms of their relationships with other entities. For example, "the black cat sitting under the table" refers to both a "black cat" entity and its relationship with another "table" entity. Understanding these relationships is essential for interpreting and grounding such natural language expressions. Most prior work focuses on either grounding entire referential expressions holistically to one region, or localizing relationships based on a fixed set of categories. In this paper we instead present a modular deep architecture capable of analyzing referential expressions into their component parts, identifying entities and relationships mentioned in the input expression and grounding them all in the scene. We call this approach Compositional Modular Networks (CMNs): a novel architecture that learns linguistic analysis and visual inference end-to-end. Our approach is built around two types of neural modules that inspect local regions and pairwise interactions between regions. We evaluate CMNs on multiple referential expression datasets, outperforming state-of-the-art approaches on all tasks.

Publications

• R. Hu, M. Rohrbach, J. Andreas, T. Darrell, K. Saenko, Modeling Relationships in Referential Expressions with Compositional Modular Networks. in CVPR, 2017
  (PDF)

@inproceedings{hu2017modeling,
  title={Modeling Relationships in Referential Expressions with Compositional Modular Networks},
  author={Hu, Ronghang and Rohrbach, Marcus and Andreas, Jacob and Darrell, Trevor and Saenko, Kate},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  year={2017}
}

Code

• Code (in TensorFlow) available at here.

Ronghang Hu, Marcus Rohrbach, Trevor Darrell

Abstract

We approach the novel problem of segmenting an image based on a natural language expression. This is different from traditional semantic segmentation over a predefined set of semantic classes, as e.g., the phrase "two men sitting on the right bench" requires segmenting only the two people on the right bench and no one standing or sitting on another bench. Previous approaches suitable for this task were limited to a fixed set of categories and/or rectangular regions. To produce pixelwise segmentation for the language expression, we propose an end-to-end trainable recurrent and convolutional network model that jointly learns to process visual and linguistic information. In our model, a recurrent LSTM network is used to encode the referential expression into a vector representation, and a fully convolutional network is used to a extract a spatial feature map from the image and output a spatial response map for the target object. We demonstrate on a benchmark dataset that our model can produce quality segmentation output from the natural language expression, and outperforms baseline methods by a large margin.

Publications

• R. Hu, M. Rohrbach, T. Darrell, Segmentation from Natural Language Expressions. in ECCV, 2016
  (PDF, Slides)

@inproceedings{hu2016segmentation,
  title={Segmentation from Natural Language Expressions},
  author={Hu, Ronghang and Rohrbach, Marcus and Darrell, Trevor},
  booktitle={Proceedings of the European Conference on Computer Vision (ECCV)},
  year={2016}
}

Data

• The training, validation and test split of ReferIt dataset used in our experiments can be downloaded here.

Code

• Code (in TensorFlow) available at here.
• Code (in Caffe, implemented by Seth Park) at here.

Ronghang Hu, Huazhe Xu, Marcus Rohrbach, Jiashi Feng, Kate Saenko, Trevor Darrell

Abstract

We address the task of natural language object retrieval, to localize a target object within a given image based on a natural language query of the object. Natural language object retrieval differs from text-based image retrieval task as it involves spatial information about objects within the scene and global scene context. To address this issue, we propose a novel Spatial Context Recurrent ConvNet (SCRC) model as scoring function on candidate boxes for object retrieval, integrating spatial configurations and global scene-level contextual information into the network. Our model processes query text, local image descriptors, spatial configurations and global context features through a recurrent network, outputs the probability of the query text conditioned on each candidate box as a score for the box, and can transfer visual-linguistic knowledge from image captioning domain to our task. Experimental results demonstrate that our method effectively utilizes both local and global information, outperforming previous baseline methods significantly on different datasets and scenarios, and can exploit large scale vision and language datasets for knowledge transfer.

Publications

• R. Hu, H. Xu, M. Rohrbach, J. Feng, K. Saenko, T. Darrell, Natural Language Object Retrieval. in CVPR, 2016
  (PDF, Slides)

@inproceedings{hu2016natural,
  title={Natural Language Object Retrieval},
  author={Hu, Ronghang and Xu, Huazhe and Rohrbach, Marcus and Feng, Jiashi and Saenko, Kate and Darrell, Trevor},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  year={2016}
}

Datasets

• The training, validation and test split of ReferIt dataset used in our experiments can be downloaded here.
• 100 top-scoring EdgeBox proposals on ReferIt dataset can be downloaded here (stored as text files per image; each row is a bounding box proposal in format, 0-indexed; can be loaded with numpy.loadtxt).
• The Kitchen dataset used in our experiment can be downloaded here.

Code

• Code (in Caffe) available at here.