VISUAL QUESTION ANSWERING

By Dr. Rhea Srinivas. 2025

Abstract - Vision-Language Pre-Training (VLP) significantly improves performance for a variety of multimodal tasks. However, existing models are often specialized in understanding or generation, which limits their versatility. Furthermore, trust in text data for large, loud web text remains the optimal approach for monitoring. To address these challenges, we propose VLX, a uniform VLP framework that distinguishes both vision languages and generation tasks. VLX introduces a new type of data optimization strategy. This strategy allows the generator to create high-quality synthetic training data, highlight the identifier noise, and allow the web to use the data records collected to more efficiently use the data records. Our framework achieves cutting-edge results with important benchmarks, including image text call (+3.1% average recall @1), visual answer questions (+2.0% accuracy), and multimodal capacitage (+2.5% cider). Additionally, VLX demonstrates the robust transferability of zero-shot transmissions to video language tasks without any additional tweaks. Publish codes, models and data records to promote future research.

Subfields: Computer Vision and Pattern Recognition, Multimodal Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Multimodal Learning

Citations in the corpus

Annotation Artifacts in Natural Language Inference Data (2018, background): “Agrawal et al. (2016); Jabri et al. (2016); Goyal et al. (2017) all showed that state-of-the-art visual question answering (Antol et al., 2015) systems leverage annotation biases in the dataset.”
From Recognition to Cognition: Visual Commonsense Reasoning (2018, neutral): “RevisitedVQA [38]: This model takes as input a query, response, and image features for the entire image, and passes the result through a multilayer perceptron, which has to classify ‘yes’ or ‘no’.”
Learning Two-Branch Neural Networks for Image-Text Matching Tasks (2017, methodology): “[9] used a softmax function to predict whether the input image and question match with”
Learning Models for Actions and Person-Object Interactions with Transfer to Question Answering (2016, background): “vity Prediction, Deep Networks, Visual Question Answering 1 Introduction The task of Visual Question Answering (VQA) has recently garnered a lot of interest with multiple datasets [1,2,3] and systems [4,5,6,7,8,9,10] being proposed. Many of these systems rely on features extracted from deep convolut”
Bottom-Up and Top-Down Attention for Image Captioning and VQA (2017, unknown): “We note that for the top-down attention component, both models use simple one-pass attention mechanisms, as opposed to the more complex schemes of recent models such as stacked, multi-headed, or bidirectional attention [52, 18, 23, 30] that could also be applied.”
Challenges and Prospects in Vision and Language Research (2019, background): “While this simplifies evaluation, it takes away a lot of the open-world difficulty from the task and can lead to inflated performance via smart guessing (Jabri et al., 2016).”
Uncovering the Temporal Context for Video Question Answering (2017, unknown): “Biases in answers Jabri et al. (2016) found that there is a strong bias in Visual7W dataset (Zhu et al. 2016). λ = 1 in Fig.”