Language understanding is essential for the navigation agent to follow instructions. We observe two kinds of issues in the instructions that can make the navigation task challenging: 1. The mentioned landmarks are not recognizable by the navigation agent due to the
different vision abilities of the instructor and the modeled agent. 2. The mentioned landmarks are applicable to multiple targets, thus not distinctive for selecting the target among
the candidate viewpoints. To deal with these issues, we design a translator module for the
navigation agent to convert the original instructions into easy-to-follow sub-instruction
representations at each step. The translator needs to focus on the recognizable and distinctive landmarks based on the agent’s visual abilities and the observed visual environment. To achieve this goal, we create a new synthetic sub-instruction dataset and design specific tasks to train the translator and the navigation agent. We evaluate our approach on Room2Room (R2R), Room4room (R4R), and Room2Room Last (R2R-Last) datasets and achieve state-of-the-art results on multiple benchmarks.
more »
« less
Zero Experience Required: Plug & Play Modular Transfer Learning for Semantic Visual Navigation
In reinforcement learning for visual navigation, it is common to develop a model for each new task, and train that model from scratch with task-specific interactions in 3D environments. However, this process is expensive; massive amounts of interactions are needed for the model to generalize well. Moreover, this process is repeated when-ever there is a change in the task type or the goal modality. We present a unified approach to visual navigation using a novel modular transfer learning model. Our model can effectively leverage its experience from one source task and apply it to multiple target tasks (e.g., ObjectNav, Room-Nav, Vi ewNav) with various goal modalities (e.g., image, sketch, audio, label). Furthermore, our model enables zero-shot experience learning, whereby it can solve the target tasks without receiving any task-specific interactive training. Our experiments on multiple photorealistic datasets and challenging tasks show that our approach learns faster, generalizes better, and outperforms SoTA models by a significant margin.
more »
« less
- Award ID(s):
- 2120430
- NSF-PAR ID:
- 10376660
- Date Published:
- Journal Name:
- IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- Page Range / eLocation ID:
- 17010 to 17020
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Few-shot knowledge graph (KG) completion task aims to perform inductive reasoning over the KG: given only a few support triplets of a new relation R (e.g., (chop, R, kitchen), (read, R, library)), the goal is to predict the query triplets of the same unseen relation R, e.g., (sleep, R, ?). Current approaches cast the problem in a meta-learning framework, where the model needs to be first jointly trained over many training few-shot tasks, each being defined by its own relation, so that learning/prediction on the target few-shot task can be effective. However, in real-world KGs, curating many training tasks is a challenging ad hoc process. We proposed Connection Subgraph Reasoner (CSR), which can make predictions for the target few-shot task directly without the need for pre-training on the human curated set of training tasks. The key to CSR is that we explicitly model a shared connection subgraph between support and query triplets, as inspired by the principle of eliminative induction. To adapt to specific KG, we design a corresponding self-supervised pretraining scheme with the objective of reconstructing automatically sampled connection subgraphs. Our pretrained model can then be directly applied to target few-shot tasks without the need for training few-shot tasks. Extensive experiments on real KGs, including NELL, FB15K-237, and ConceptNet, demonstrate the effectiveness of our framework: we have shown that even a learning-free implementation of CSR can already perform competitively to existing methods on target few-shot tasks; with pretraining, CSR can achieve significant gains of up to 52% on the more challenging inductive few-shot tasks where the entities are also unseen during (pre)training.more » « less
-
more » « less
Abstract Investigations into how individual neurons encode behavioral variables of interest have revealed specific representations in single neurons, such as place and object cells, as well as a wide range of cells with conjunctive encodings or mixed selectivity. However, as most experiments examine neural activity within individual tasks, it is currently unclear if and how neural representations change across different task contexts. Within this discussion, the medial temporal lobe is particularly salient, as it is known to be important for multiple behaviors including spatial navigation and memory, however the relationship between these functions is currently unclear. Here, to investigate how representations in single neurons vary across different task contexts in the medial temporal lobe, we collected and analyzed single‐neuron activity from human participants as they completed a paired‐task session consisting of a passive‐viewing visual working memory and a spatial navigation and memory task. Five patients contributed 22 paired‐task sessions, which were spike sorted together to allow for the same putative single neurons to be compared between the different tasks. Within each task, we replicated concept‐related activations in the working memory task, as well as target‐location and serial‐position responsive cells in the navigation task. When comparing neuronal activity between tasks, we first established that a significant number of neurons maintained the same kind of representation, responding to stimuli presentations across tasks. Further, we found cells that changed the nature of their representation across tasks, including a significant number of cells that were stimulus responsive in the working memory task that responded to serial position in the spatial task. Overall, our results support a flexible encoding of multiple, distinct aspects of different tasks by single neurons in the human medial temporal lobe, whereby some individual neurons change the nature of their feature coding between task contexts.
-
We address the challenge of getting efficient yet accurate recognition systems with limited labels. While recognition models improve with model size and amount of data, many specialized applications of computer vision have severe resource constraints both during training and inference. Transfer learning is an effective solution for training with few labels, however often at the expense of a compu- tationally costly fine-tuning of large base models. We propose to mitigate this unpleasant trade-off between compute and accuracy via semi-supervised cross- domain distillation from a set of diverse source models. Initially, we show how to use task similarity metrics to select a single suitable source model to distill from, and that a good selection process is imperative for good downstream performance of a target model. We dub this approach DISTILLNEAREST. Though effective, DISTILLNEAREST assumes a single source model matches the target task, which is not always the case. To alleviate this, we propose a weighted multi-source distilla- tion method to distill multiple source models trained on different domains weighted by their relevance for the target task into a single efficient model (named DISTILL- WEIGHTED). Our methods need no access to source data, and merely need features and pseudo-labels of the source models. When the goal is accurate recognition under computational constraints, both DISTILLNEAREST and DISTILLWEIGHTED approaches outperform both transfer learning from strong ImageNet initializations as well as state-of-the-art semi-supervised techniques such as FixMatch. Averaged over 8 diverse target tasks our multi-source method outperforms the baselines by 5.6%-points and 4.5%-points, respectively.more » « less
-
The advances in deep reinforcement learning re- cently revived interest in data-driven learning based approaches to navigation. In this paper we propose to learn viewpoint invariant and target invariant visual servoing for local mobile robot navigation; given an initial view and the goal view or an image of a target, we train deep convolutional network controller to reach the desired goal. We present a new architecture for this task which rests on the ability of establishing correspondences between the initial and goal view and novel reward structure motivated by the traditional feedback control error. The advantage of the proposed model is that it does not require calibration and depth information and achieves robust visual servoing in a variety of environments and targets without any parameter fine tuning. We present comprehensive evaluation of the approach and comparison with other deep learning architectures as well as classical visual servoing methods in visually realistic simulation environment [1]. The presented model overcomes the brittleness of classical visual servoing based methods and achieves significantly higher generalization capability compared to the previous learning approaches.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/CVPR52688.2022.01652
