Development of a Multimodal AI Visual Chatbot Using Python for Enhanced Human-Computer Interaction

From this perspective, the proposed Visual Chatbot v2.0 showcases an efficient combination of NLP and computer vision technologies to create a human-computer interaction tool with multiple capabilities. First, unlike standard chatbots that are limited to text-based dialogue, the designed solution allows the analysis of images and videos and can generate context-aware responses. With the use of Python programming and Gemini vision-language model, the chatbot is able to respond meaningfully and instantly. In addition, its implementation involves such features as media preprocessing, multi-frame video analysis, conversation memory, and graphical user interface (GUI), making the proposed solution applicable in practice.  

[1]   G. Zhao et al., “NavGemini: A Multi-Modal LLM Agent for Vision-and-Language Tasks,” Springer, 2026. [2]    Y. Li et al., “Mini-Gemini: Enhancing Multi-Modality Vision Language Models,” arXiv preprint arXiv:2403.18814, 2024. [3]   B. R. Kim et al., “Multimodal Large Language Models in Real-World Applications,” ScienceDirect, 2025. [4]    F. Lin, “Vision Language Models: A Survey of 26K Papers,” arXiv preprint, 2025. [5]   L. Yan et al., “A Comprehensive Survey of Multimodal LLMs,” OpenReview, 2025. [6]  G. Rubegni et al., “Applications of Large Language Models in Vision-Based Systems,” PMC, 2026. [7]     Google DeepMind, “Gemini: Multimodal AI Model Capabilities,” 2025. [8]      Google, “Gemini Language Model Overview and Evolution,” 2025. [9]  OpenAI, “GPT-5: A Multimodal Large Language Model,” 2025. [10]   A. Pal and M. Sankarasubbu, “Evaluation of Gemini in Multimodal Medical Tasks,” arXiv, 2024.