paper-with-me

홈 › Papers

WinoViz: Probing Visual Properties of Objects Under Different States

2024-02-21 · Woojeong Jin, Tejas Srinivasan, Jesse Thomason, Xiang Ren

Humans perceive and comprehend different visual properties of an object based on specific contexts. For instance, we know that a banana turns brown `when it becomes rotten,'' whereas it appears green `when it is unripe.'' Previous studies on probing visual commonsense knowledge have primarily focused on examining language models' understanding of typical properties (e.g., colors and shapes) of objects. We present WinoViz, a text-only evaluation dataset, consisting of 1,380 examples that probe the reasoning abilities of language models regarding variant visual properties of objects under different contexts or states. Our task is challenging since it requires pragmatic reasoning (finding intended meanings) and visual knowledge reasoning. We also present multi-hop data, a more challenging version of our data, which requires multi-step reasoning chains to solve our task. In our experimental analysis, our findings are: a) Large language models such as GPT-4 demonstrate effective performance, but when it comes to multi-hop data, their performance is significantly degraded. b) Large models perform well on pragmatic reasoning, but visual knowledge reasoning is a bottleneck in our task. c) Vision-language models outperform their language-model counterparts. d) A model with machine-generated images performs poorly in our task. This is due to the poor quality of the generated images.

📄 PDF Abstract BibTeX arXiv:2402.13584

Code (0)

등록된 구현이 없습니다.

Tasks

Language Modelling

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
Adam 설명 없음
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Multi-Head Attention 설명 없음
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…

Similar Papers 제목 키워드 기반

Probing Multimodal Embeddings for Linguistic Properties: the Visual-Semantic Case

2021-02-22 · COLING 2020 8 · Adam Dahlgren Lindström, Suna Bensch, Johanna Björklund, Frank Drewes

Semantic embeddings have advanced the state of the art for countless natural language processing tasks, and various extensions to multimodal domains, such as visual-semantic embeddings, have been proposed. While the powe…

Probing Emergent Semantics in Predictive Agents via Question Answering

2020-06-01 · ICML 2020 1 · Abhishek Das, Federico Carnevale, Hamza Merzic, Laura Rimell 외

Recent work has shown how predictive modeling can endow agents with rich knowledge of their surroundings, improving their ability to act in complex environments. We propose question-answering as a general paradigm to dec…

DecoderQuestion Answering

Understanding Physical Properties of Unseen Deformable Objects by Leveraging Large Language Models and Robot Actions

2025-06-04 · Changmin Park, Beomjoon Lee, Haechan Jung, Haejin Jung 외

In this paper, we consider the problem of understanding the physical properties of unseen objects through interactions between the objects and a robot. Handling unseen objects with special properties such as deformabilit…

Motion PlanningTask and Motion PlanningTask Planning

Probing the Role of Positional Information in Vision-Language Models

2022-01-16 · ACL ARR January 2022 1 · Anonymous

In most Vision-Language models (VL) the understanding of the image structure is enabled by injecting the position information (PI) about objects in the image. In our case study of LXMERT, a state-of-the-art VL model, we …

Contrastive LearningImage-text matchingObject LocalizationPosition+4

Probing the Role of Positional Information in Vision-Language Models

2023-05-17 · Findings (NAACL) 2022 7 · Philipp J. Rösch, Jindřich Libovický

In most Vision-Language models (VL), the understanding of the image structure is enabled by injecting the position information (PI) about objects in the image. In our case study of LXMERT, a state-of-the-art VL model, we…

Contrastive LearningImage-text matchingObject LocalizationPosition+3