Revisiting Image Captioning Training Paradigm via Direct CLIP-based Optimization
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_ | 1866909297026269184 |
|---|---|
| author | Moratelli, Nicholas Caffagni, Davide Cornia, Marcella Baraldi, Lorenzo Cucchiara, Rita |
| author_facet | Moratelli, Nicholas Caffagni, Davide Cornia, Marcella Baraldi, Lorenzo Cucchiara, Rita |
| contents | The conventional training approach for image captioning involves pre-training a network using teacher forcing and subsequent fine-tuning with Self-Critical Sequence Training to maximize hand-crafted captioning metrics. However, when attempting to optimize modern and higher-quality metrics like CLIP-Score and PAC-Score, this training method often encounters instability and fails to acquire the genuine descriptive capabilities needed to produce fluent and informative captions. In this paper, we propose a new training paradigm termed Direct CLIP-Based Optimization (DiCO). Our approach jointly learns and optimizes a reward model that is distilled from a learnable captioning evaluator with high human correlation. This is done by solving a weighted classification problem directly inside the captioner. At the same time, DiCO prevents divergence from the original model, ensuring that fluency is maintained. DiCO not only exhibits improved stability and enhanced quality in the generated captions but also aligns more closely with human preferences compared to existing methods, especially in modern metrics. Additionally, it maintains competitive performance in traditional metrics. Our source code and trained models are publicly available at https://github.com/aimagelab/DiCO. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_14547 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Revisiting Image Captioning Training Paradigm via Direct CLIP-based Optimization Moratelli, Nicholas Caffagni, Davide Cornia, Marcella Baraldi, Lorenzo Cucchiara, Rita Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Multimedia The conventional training approach for image captioning involves pre-training a network using teacher forcing and subsequent fine-tuning with Self-Critical Sequence Training to maximize hand-crafted captioning metrics. However, when attempting to optimize modern and higher-quality metrics like CLIP-Score and PAC-Score, this training method often encounters instability and fails to acquire the genuine descriptive capabilities needed to produce fluent and informative captions. In this paper, we propose a new training paradigm termed Direct CLIP-Based Optimization (DiCO). Our approach jointly learns and optimizes a reward model that is distilled from a learnable captioning evaluator with high human correlation. This is done by solving a weighted classification problem directly inside the captioner. At the same time, DiCO prevents divergence from the original model, ensuring that fluency is maintained. DiCO not only exhibits improved stability and enhanced quality in the generated captions but also aligns more closely with human preferences compared to existing methods, especially in modern metrics. Additionally, it maintains competitive performance in traditional metrics. Our source code and trained models are publicly available at https://github.com/aimagelab/DiCO. |
| title | Revisiting Image Captioning Training Paradigm via Direct CLIP-based Optimization |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Multimedia |
| url | https://arxiv.org/abs/2408.14547 |