Biomedical Visual Instruction Tuning with Clinician Preference Alignment
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911956613464064 |
|---|---|
| author | Cui, Hejie Mao, Lingjun Liang, Xin Zhang, Jieyu Ren, Hui Li, Quanzheng Li, Xiang Yang, Carl |
| author_facet | Cui, Hejie Mao, Lingjun Liang, Xin Zhang, Jieyu Ren, Hui Li, Quanzheng Li, Xiang Yang, Carl |
| contents | Recent advancements in multimodal foundation models have showcased impressive capabilities in understanding and reasoning with visual and textual information. Adapting these foundation models trained for general usage to specialized domains like biomedicine requires large-scale domain-specific instruction datasets. While existing works have explored curating such datasets automatically, the resultant datasets are not explicitly aligned with domain expertise. In this work, we propose a data-centric framework, Biomedical Visual Instruction Tuning with Clinician Preference Alignment (BioMed-VITAL), that incorporates clinician preferences into both stages of generating and selecting instruction data for tuning biomedical multimodal foundation models. First, during the generation stage, we prompt the GPT-4V generator with a diverse set of clinician-selected demonstrations for preference-aligned data candidate generation. Then, during the selection phase, we train a separate selection model, which explicitly distills clinician and policy-guided model preferences into a rating function to select high-quality data for medical instruction tuning. Results show that the model tuned with the instruction-following data from our method demonstrates a significant improvement in open visual chat (18.5% relatively) and medical VQA (win rate up to 81.73%). Our instruction-following data and models are available at BioMed-VITAL.github.io. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_13173 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Biomedical Visual Instruction Tuning with Clinician Preference Alignment Cui, Hejie Mao, Lingjun Liang, Xin Zhang, Jieyu Ren, Hui Li, Quanzheng Li, Xiang Yang, Carl Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Machine Learning 68T50, 68T45, 68T37, 68T05, 68T07, 68T09, I.2.7; I.2.6; I.2.10 Recent advancements in multimodal foundation models have showcased impressive capabilities in understanding and reasoning with visual and textual information. Adapting these foundation models trained for general usage to specialized domains like biomedicine requires large-scale domain-specific instruction datasets. While existing works have explored curating such datasets automatically, the resultant datasets are not explicitly aligned with domain expertise. In this work, we propose a data-centric framework, Biomedical Visual Instruction Tuning with Clinician Preference Alignment (BioMed-VITAL), that incorporates clinician preferences into both stages of generating and selecting instruction data for tuning biomedical multimodal foundation models. First, during the generation stage, we prompt the GPT-4V generator with a diverse set of clinician-selected demonstrations for preference-aligned data candidate generation. Then, during the selection phase, we train a separate selection model, which explicitly distills clinician and policy-guided model preferences into a rating function to select high-quality data for medical instruction tuning. Results show that the model tuned with the instruction-following data from our method demonstrates a significant improvement in open visual chat (18.5% relatively) and medical VQA (win rate up to 81.73%). Our instruction-following data and models are available at BioMed-VITAL.github.io. |
| title | Biomedical Visual Instruction Tuning with Clinician Preference Alignment |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Machine Learning 68T50, 68T45, 68T37, 68T05, 68T07, 68T09, I.2.7; I.2.6; I.2.10 |
| url | https://arxiv.org/abs/2406.13173 |