AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Prompt Tuning of Deep Neural Networks for Speaker-Adaptive Visual Speech Recognition

Minsu Kim, Hyung-Il Kim, Yong Man Ro

TPAMI 2024 · Vision-language and multimodal › CLIP and zero-shot · approach: Prompt learning for CLIP · 34 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
tune speaker prompts on 5 min of adaptation datapre-trained VSR · GRIDmean WER: 12.04 -> 3.80Improved
tune speaker prompts on 5 min of adaptation datapre-trained VSR · LRW-IDmean accuracy: 87.54 -> 89.99Improved
add concatenation prompt to Add and Pad prompts (ablation)CNN prompts · GRIDmean WER (5 min): 4.99 -> 4.04 (Pad vs Add+Pad); 4.17 -> 3.80 (Add vs Add+Pad+Cat)Improved
use padding prompt instead of addition promptsingle prompt, deep CNN · LRW-IDmean accuracy (5 min): 88.03 -> 89.62Improved
use addition prompt instead of padding promptsingle prompt, shallow CNN · GRIDmean WER (5 min): 4.99 -> 4.17Improved
increase adaptation data from 1 to 5 minAdd+Pad+Cat · GRIDmean WER: 5.53 -> 3.80Improved
add padding-prompt layers (N_l 5 -> 17) (ablation)padding prompt · LRW-IDaccuracy (5 min): 88.55 -> 89.62Improved
increase concatenation prompt length N_p 1 -> 5 (ablation)concatenation prompt · LRW-IDaccuracy (1 and 3 min): 88.61 -> 88.92 (3 min); 88.27 -> 88.28 (1 min)Improved
use prompt tuning instead of full finetuning (small data)Pad+Cat prompt · LRW-IDaccuracy (5 min): 89.91 -> 89.99Improved
use prompt tuning instead of full finetuningAdd+Pad+Cat prompt · GRIDWER (5 min): 3.60 -> 3.80Hurt
use prompt tuning instead of full finetuning with 100% adaptation dataAdd+Pad+Cat prompt · GRIDWER: 2.34 -> 3.13Hurt
use prompt tuning instead of Speaker codeAdd+Cat prompt · GRIDWER (5 min): 4.68 -> 4.00Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy