Video captioning is a task of describing the visual scene of a given video in natural language. There have been several lines of research focused on developing large-scale models in a transfer learning paradigm, with major challenge being the tradeoff between scalability and performance in limited environments. To address this problem, we propose a simple yet effective encoder–decoder-based video captioning model integrating transformers and CLIP, both of which are widely adopted in the vision and language domains, together with appropriate temporal feature embedding modules. Taking this proposal a step further, we also address the challenge of human-interactive video captioning, where the captions are tailored to specific information desired by humans. To design a human-interactive environment, we assume that a human offers an object or action in the video as a short prompt; in turn, the system then provides a detailed explanation regarding the prompt. We embed human prompts within an LSTM-based prompt encoder and leverage soft prompting to tune the model effectively. We extensively evaluated our model on benchmark datasets, demonstrating comparable results, particularly on the MSR-VTT dataset, where we achieve state-of-the-art performance with 4% improvement. In addition, we also show potential for human-interactive video captioning through quantitative and qualitative analysis.

Deep neural networks have been successfully adopted to diverse domains including pathology classification based on medical images. However, large-scale and high-quality data to train powerful neural networks are rare in the medical domain as the labeling must be done by qualified experts. Researchers recently tackled this problem with some success by taking advantage of models pre-trained on large-scale general domain data. Specifically, researchers took contrastive image-text encoders (e.g., CLIP) and fine-tuned it with chest X-ray images and paired reports to perform zero-shot pathology classification, thus completely removing the need for pathology-annotated images to train a classification model. Existing studies, however, fine-tuned the pre-trained model with the same contrastive learning objective, and failed to exploit the multi-labeled nature of medical image-report pairs. In this paper, we propose a new fine-tuning strategy based on sentence sampling and positive pair loss relaxation for improving the downstream zero-shot pathology classification performance, which can be applied to any pre-trained contrastive image-text encoders. Our method consistently showed dramatically improved zero-shot pathology classification performance on four different chest X-ray datasets and 3 different pre-trained models (5.77% average AUROC increase). In particular, fine-tuning CLIP with our method showed much comparable or marginally outperformed to board-certified radiologists (0.619 vs 0.625 in F1 score and 0.530 vs 0.544 in MCC) in zero-shot classification of five prominent diseases from the CheXpert dataset.

Pre-training vision-language models with contrastive objectives has shown promising results that are both scalable to large uncurated datasets and transferable to many downstream applications. Some following works have targeted to improve data efficiency by adding self-supervision terms, but inter-domain (image-text) contrastive loss and intra-domain (image-image) contrastive loss are defined on individual spaces in those works, so many feasible combinations of supervision are overlooked. To overcome this issue, we propose UniCLIP, a Unified framework for Contrastive Language-Image Pre-training. UniCLIP integrates the contrastive loss of both inter-domain pairs and intra-domain pairs into a single universal space. The discrepancies that occur when integrating contrastive loss between different domains are resolved by the three key components of UniCLIP: (1) augmentation-aware feature embedding, (2) MP-NCE loss, and (3) domain dependent similarity measure. UniCLIP outperforms previous vision-language pre-training methods on various single- and multi-modality downstream tasks. In our experiments, we show that each component that comprises UniCLIP contributes well to the final performance.

In this report, we introduce NICE (New frontiers for zero-shot Image Captioning Evaluation) project and share the results and outcomes of 2023 challenge. This project is designed to challenge the computer vision community to develop robust image captioning models that advance the state-of-the-art both in terms of accuracy and fairness. Through the challenge, the image captioning models were tested using a new evaluation dataset that includes a large variety of visual concepts from many domains. There was no specific training data provided for the challenge, and therefore the challenge entries were required to adapt to new types of image descriptions that had not been seen during training. This report includes information on the newly proposed NICE dataset, evaluation methods, challenge results, and technical details of top-ranking entries. We expect that the outcomes of the challenge will contribute to the improvement of AI models on various vision-language tasks.