An official website of the United States government
Internet image collections containing photos captured by crowds of photographers show promise for enabling digital exploration of large‐scale tourist landmarks. However, prior works focus primarily on geometric reconstruction and visualization, neglecting the key role of language in providing a semantic interface for navigation and fine‐grained understanding. In more constrained 3D domains, recent methods have leveraged modern vision‐and‐language models as a strong prior of 2D visual semantics. While these models display an excellent understanding of broad visual semantics, they struggle with unconstrained photo collections depicting such tourist landmarks, as they lack expert knowledge of the architectural domain and fail to exploit the geometric consistency of images capturing multiple views of such scenes. In this work, we present a localization system that connects neural representations of scenes depicting large‐scale landmarks with text describing a semantic region within the scene, by harnessing the power of SOTA vision‐and‐language models with adaptations for understanding landmark scene semantics. To bolster such models with fine‐grained knowledge, we leverage large‐scale Internet data containing images of similar landmarks along with weakly‐related textual information. Our approach is built upon the premise that images physically grounded in space can provide a powerful supervision signal for localizing new concepts, whose semantics may be unlocked from Internet textual metadata with large language models. We use correspondences between views of scenes to bootstrap spatial understanding of these semantics, providing guidance for 3D‐compatible segmentation that ultimately lifts to a volumetric scene representation. To evaluate our method, we present a new benchmark dataset containing large‐scale scenes with ground‐truth segmentations for multiple semantic concepts. Our results show that HaLo‐NeRF can accurately localize a variety of semantic concepts related to architectural landmarks, surpassing the results of other 3D models as well as strong 2D segmentation baselines. Our code and data are publicly available at https://tau‐vailab.github.io/HaLo‐NeRF/
more »
« less
Rel3D: A Minimally Contrastive Benchmark for Grounding Spatial Relations in 3D
Understanding spatial relations (e.g., laptop on table) in visual input is important for both humans and robots. Existing datasets are insufficient as they lack large-scale, high-quality 3D ground truth information, which is critical for learning spatial relations. In this paper, we fill this gap by constructing Rel3D: the first large-scale, human-annotated dataset for grounding spatial relations in 3D. Rel3D enables quantifying the effectiveness of 3D information in predicting spatial relations on large-scale human data. Moreover, we propose minimally contrastive data collection---a novel crowdsourcing method for reducing dataset bias. The 3D scenes in our dataset come in minimally contrastive pairs: two scenes in a pair are almost identical, but a spatial relation holds in one and fails in the other. We empirically validate that minimally contrastive examples can diagnose issues with current relation detection models as well as lead to sample-efficient training. Code and data are available at https://github.com/princeton-vl/Rel3D.
more »
« less
- Award ID(s):
- 1734266
- PAR ID:
- 10202375
- Date Published:
- Journal Name:
- Advances in Neural Information Processing Systems 33 pre-proceedings (NeurIPS)
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Humans subconsciously engage in geospatial reasoning when reading articles. We recognize place names and their spatial relations in text and mentally associate them with their physical locations on Earth. Although pretrained language models can mimic this cognitive process using linguistic context, they do not utilize valuable geospatial information in large, widely available geographical databases, e.g., OpenStreetMap. This paper introduces GeoLM, a geospatially grounded language model that enhances the understanding of geo-entities in natural language. GeoLM leverages geo-entity mentions as anchors to connect linguistic information in text corpora with geospatial information extracted from geographical databases. GeoLM connects the two types of context through contrastive learning and masked language modeling. It also incorporates a spatial coordinate embedding mechanism to encode distance and direction relations to capture geospatial context. In the experiment, we demonstrate that GeoLM exhibits promising capabilities in supporting toponym recognition, toponym linking, relation extraction, and geo-entity typing, which bridge the gap between natural language processing and geospatial sciences. The code is publicly available at https://github.com/knowledge-computing/geolm.more » « less
-
Abstract Objects and places are foundational spatial domains represented in human symbolic expressions, like drawings, which show a prioritization of depicting small-scale object-shape information over the large-scale navigable place information in which objects are situated. Is there a similar object-over-place bias in language? Across six experiments, adults and 3- to 4-year-old children were asked either to extend a novel noun in a labeling phrase, to extend a novel noun in a prepositional phrase, or to simply match pictures. To dissociate specific object and place information from more general figure and ground information, participants either saw scenes with both place information (a room) and object information (a block in the room), or scenes with two kinds of object information that matched the figure-ground relations of the room and block by presenting an open container with a smaller block inside. While adults showed a specific object-over-place bias in both extending novel noun labels and matching, they did not show this bias in extending novel nouns following prepositions. Young children showed this bias in extending novel noun labels only. Spatial domains may thus confer specific and foundational biases for word learning that may change through development in a way that is similar to that of other word-learning biases about objects, like the shape bias. These results expand the symbolic scope of prior studies on object biases in drawing to object biases in language, and they expand the spatial domains of prior studies characterizing the language of objects and places.more » « less
-
null (Ed.)Arguably one of the top success stories of deep learning is transfer learning. The finding that pre-training a network on a rich source set (e.g., ImageNet) can help boost performance once fine-tuned on a usually much smaller target set, has been instrumental to many applications in language and vision. Yet, very little is known about its usefulness in 3D point cloud understanding. We see this as an opportunity considering the effort required for annotating data in 3D. In this work, we aim at facilitating research on 3D representation learning. Different from previous works, we focus on high-level scene understanding tasks. To this end, we select a suit of diverse datasets and tasks to measure the effect of unsupervised pre-training on a large source set of 3D scenes. Our findings are extremely encouraging: using a unified triplet of architecture, source dataset, and contrastive loss for pre-training, we achieve improvement over recent best results in segmentation and detection across 6 different benchmarks for indoor and outdoor, real and synthetic datasets – demonstrating that the learned representation can generalize across domains. Furthermore, the improvement was similar to supervised pre-training, suggesting that future efforts should favor scaling data collection over more detailed annotation. We hope these findings will encourage more research on unsupervised pretext task design for 3D deep learning.more » « less
-
Knowledge graphs (KGs) are of great importance in various artificial intelligence systems, such as question answering, relation extraction, and recommendation. Nevertheless, most real-world KGs are highly incomplete, with many missing relations between entities. To discover new triples (i.e., head entity, relation, tail entity), many KG completion algorithms have been proposed in recent years. However, a vast majority of existing studies often require a large number of training triples for each relation, which contradicts the fact that the frequency distribution of relations in KGs often follows a long tail distribution, meaning a majority of relations have only very few triples. Meanwhile, since most existing large-scale KGs are constructed automatically by extracting information from crowd-sourcing data using heuristic algorithms, plenty of errors could be inevitably incorporated due to the lack of human verification, which greatly reduces the performance for KG completion. To tackle the aforementioned issues, in this paper, we study a novel problem of error-aware few-shot KG completion and present a principled KG completion framework REFORM. Specifically, we formulate the problem under the few-shot learning framework, and our goal is to accumulate meta-knowledge across different meta-tasks and generalize the accumulated knowledge to the meta-test task for error-aware few-shot KG completion. To address the associated challenges resulting from insufficient training samples and inevitable errors, we propose three essential modules neighbor encoder, cross-relation aggregation, and error mitigation in each meta-task. Extensive experiments on three widely used KG datasets demonstrate the superiority of the proposed framework REFORM over competitive baseline methods.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Conference Paper:
- The DOI is not currently available.
