Zobrazeno 1 - 10
of 25
pro vyhledávání: '"Kreiss, Elisa"'
Although CLIPScore is a powerful generic metric that captures the similarity between a text and an image, it fails to distinguish between a caption that is meant to complement the information in an image and a description that is meant to replace an
Externí odkaz:
http://arxiv.org/abs/2406.09458
Current visual question answering (VQA) models tend to be trained and evaluated on image-question pairs in isolation. However, the questions people ask are dependent on their informational needs and prior knowledge about the image content. To evaluat
Externí odkaz:
http://arxiv.org/abs/2402.15002
Referenceless metrics (e.g., CLIPScore) use pretrained vision--language models to assess image descriptions directly without costly ground-truth reference texts. Such methods can facilitate rapid progress, but only if they truly align with human pref
Externí odkaz:
http://arxiv.org/abs/2309.11710
Visual question answering (VQA) has the potential to make the Internet more accessible in an interactive way, allowing people who cannot see images to ask questions about them. However, multiple studies have shown that people who are blind or have lo
Externí odkaz:
http://arxiv.org/abs/2307.15745
Autor:
Kreiss, Elisa, Srinivasan, Krishna, Piccardi, Tiziano, Hermosillo, Jesus Adolfo, Bennett, Cynthia, Bernstein, Michael S., Morris, Meredith Ringel, Potts, Christopher
We make a first attempt to characterize image accessibility on Wikipedia across languages, present new experimental results that can inform efforts to assess description quality, and offer some strategies to improve Wikipedia's image accessibility.
Externí odkaz:
http://arxiv.org/abs/2305.09038
Autor:
Kreiss, Elisa, Bennett, Cynthia, Hooshmand, Shayan, Zelikman, Eric, Morris, Meredith Ringel, Potts, Christopher
Few images on the Web receive alt-text descriptions that would make them accessible to blind and low vision (BLV) users. Image-based NLG systems have progressed to the point where they can begin to address this persistent societal problem, but these
Externí odkaz:
http://arxiv.org/abs/2205.10646
Speakers' referential expressions often depart from communicative ideals in ways that help illuminate the nature of pragmatic language use. Patterns of overmodification, in which a speaker uses a modifier that is redundant given their communicative g
Externí odkaz:
http://arxiv.org/abs/2205.09172
Autor:
Wu, Zhengxuan, Geiger, Atticus, Rozner, Josh, Kreiss, Elisa, Lu, Hanson, Icard, Thomas, Potts, Christopher, Goodman, Noah D.
Publikováno v:
NAACL 2022
Distillation efforts have led to language models that are more compact and efficient without serious drops in performance. The standard approach to distillation trains a student model against two objectives: a task-specific objective (e.g., language
Externí odkaz:
http://arxiv.org/abs/2112.02505
Autor:
Geiger, Atticus, Wu, Zhengxuan, Lu, Hanson, Rozner, Josh, Kreiss, Elisa, Icard, Thomas, Goodman, Noah D., Potts, Christopher
In many areas, we have well-founded insights about causal structure that would be useful to bring into our trained models while still allowing them to learn in a data-driven fashion. To achieve this, we present the new method of interchange intervent
Externí odkaz:
http://arxiv.org/abs/2112.00826
Publikováno v:
NeurIPS 2021 (Dataset and Benchmarks Track)
The ability to compositionally map language to referents, relations, and actions is an essential component of language understanding. The recent gSCAN dataset (Ruis et al. 2020, NeurIPS) is an inspiring attempt to assess the capacity of models to lea
Externí odkaz:
http://arxiv.org/abs/2109.08994