Recently, reference-free metrics such as CLIPScore (Hessel et al., 2021), UMIC (Lee et al., 2021), and PAC-S (Sarto et al., 2023) have been proposed for automatic reference-free evaluation of image captions. Our focus lies in evaluating the robustness of these metrics in scenarios that require distinguishing between two captions with high lexical overlap but very different meanings. Our findings reveal that despite their high correlation with human judgments, CLIPScore, UMIC, and PAC-S struggle to identify fine-grained errors. While all metrics exhibit strong sensitivity to visual grounding errors, their sensitivity to caption implausibility errors is limited. Furthermore, we found that all metrics are sensitive to variations in the size of image-relevant objects mentioned in the caption, while CLIPScore and PAC-S are also sensitive to the number of mentions of image-relevant objects in the caption. Regarding linguistic aspects of a caption, all metrics show weak comprehension of negation, and CLIPScore and PAC-S are insensitive to the structure of the caption to a great extent. We hope our findings will guide further improvements in reference-free evaluation of image captioning.

Vision-and-language (V&L) models pretrained on large-scale multimodal data have demonstrated strong performance on various tasks such as image captioning and visual question answering (VQA). The quality of such models is commonly assessed by measuring their performance on unseen data that typically comes from the same distribution as the training data. However, when evaluated under out-of-distribution (out-of-dataset) settings for VQA, we observe that these models exhibit poor generalization. We comprehensively evaluate two pretrained V&L models under different settings (i.e. classification and open-ended text generation) by conducting cross-dataset evaluations. We find that these models tend to learn to solve the benchmark, rather than learning the high-level skills required by the VQA task. We also find that in most cases generative models are less susceptible to shifts in data distribution compared to discriminative ones, and that multimodal pretraining is generally helpful for OOD generalization. Finally, we revisit assumptions underlying the use of automatic VQA evaluation metrics, and empirically show that their stringent nature repeatedly penalizes models for correct responses.

Multi-modal open-domain question answering typically requires evidence retrieval from databases across diverse modalities, such as images, tables, passages, etc. Even Large Language Models (LLMs) like GPT-4 fall short in this task. To enable LLMs to tackle the task in a zero-shot manner, we introduce MoqaGPT, a straightforward and flexible framework. Using a divide-and-conquer strategy that bypasses intricate multi-modality ranking, our framework can accommodate new modalities and seamlessly transition to new models for the task. Built upon LLMs, MoqaGPT retrieves and extracts answers from each modality separately, then fuses this multi-modal information using LLMs to produce a final answer. Our methodology boosts performance on the MMCoQA dataset, improving F1 by +37.91 points and EM by +34.07 points over the supervised baseline. On the MultiModalQA dataset, MoqaGPT surpasses the zero-shot baseline, improving F1 by 9.5 points and EM by 10.1 points, and significantly closes the gap with supervised methods. Our codebase is available at https://github.com/lezhang7/MOQAGPT.

While pretraining on large-scale image–text data from the Web has facilitated rapid progress on many vision-and-language (V&L) tasks, recent work has demonstrated that pretrained models lack “fine-grained” understanding, such as the ability to recognise relationships, verbs, and numbers in images. This has resulted in an increased interest in the community to either develop new benchmarks or models for such capabilities. To better understand and quantify progress in this direction, we investigate four competitive V&L models on four fine-grained benchmarks. Through our analysis, we find that X-VLM (Zeng et al., 2022) consistently outperforms other baselines, and that modelling innovations can impact performance more than scaling Web data, which even degrades performance sometimes. Through a deeper investigation of X-VLM, we highlight the importance of both novel losses and rich data sources for learning fine-grained skills. Finally, we inspect training dynamics, and discover that for some tasks, performance peaks early in training or significantly fluctuates, never converging.

Large pre-trained models have proved to be remarkable zero- and (prompt-based) few-shot learners in unimodal vision and language tasks. We propose MAPL, a simple and parameter-efficient method that reuses frozen pre-trained unimodal models and leverages their strong generalization capabilities in multimodal vision-language (VL) settings. MAPL learns a lightweight mapping between the representation spaces of unimodal models using aligned image-text data, and can generalize to unseen VL tasks from just a few in-context examples. The small number of trainable parameters makes MAPL effective at low-data and in-domain learning. Moreover, MAPL’s modularity enables easy extension to other pre-trained models. Extensive experiments on several visual question answering and image captioning benchmarks show that MAPL achieves superior or competitive performance compared to similar methods while training orders of magnitude fewer parameters. MAPL can be trained in just a few hours using modest computational resources and public datasets. We release our code and pre-trained model weights at https://github.com/oscmansan/mapl.

In the last few years, there has been an increased interest in building multimodal (vision-language) models that are pretrained on larger but noisier datasets where the two modalities (e.g., image and text) loosely correspond to each other (e.g., Lu et al., 2019; Radford et al., 2021). Given a task (such as visual question answering), these models are then often fine-tuned on task-specific supervised datasets. (e.g., Lu et al., 2019; Chen et al.,2020; Tan and Bansal, 2019; Li et al., 2020a,b). In addition to the larger pretraining datasets, the transformer architecture (Vaswani et al., 2017) and in particular self-attention applied to two modalities are responsible for the impressive performance of the recent pretrained models on downstream tasks (Hendricks et al., 2021). In this tutorial, we focus on recent vision-language pretraining paradigms. Our goal is to first provide the background on image–language datasets, benchmarks, and modeling innovations before the multimodal pretraining area. Next we discuss the different family of models used for vision-language pretraining, highlighting their strengths and shortcomings. Finally, we discuss the limits of vision-language pretraining through statistical learning, and the need for alternative approaches such as causal representation learning.