Machine learningDeep learning / NLP / CV

बहुविध छवि वर्गीकरण

बहुविध छवि वर्गीकरण (Multimodal image classification) मानक दृश्य वर्गीकरण का विस्तार करता है, जिसमें छवि विशेषताओं के साथ-साथ अतिरिक्त विधाओं - जैसे पाठ्य कैप्शन, ऑडियो, या संरचित मेटाडेटा - को शामिल किया जाता है। अलग-अलग एनकोडर प्रत्येक विधा को संसाधित करते हैं, उनके निरूपण (representations) को संयोजित (fused) किया जाता है, और एक संयुक्त क्लासिफायर लक्ष्य लेबल निर्दिष्ट करता है। CLIP जैसे मॉडल प्रदर्शित करते हैं कि छवि-पाठ संरेखण (image–text alignment) बड़े पैमाने पर शून्य-शॉट (zero-shot) और कुछ-शॉट (few-shot) छवि वर्गीकरण को सक्षम बनाता है।

MethodMind में खोलेंजल्द हीवीडियोजल्द हीDownload slides

पूरी विधि पढ़ें

केवल सदस्यों के लिए

यह खंड पढ़ने के लिए निःशुल्क खाते से साइन इन करें।

साइन इन करें

Method map

The neighbourhood of related methods — select a node to explore.

स्रोत

  1. Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., ... & Sutskever, I. (2021). Learning transferable visual models from natural language supervision. Proceedings of the 38th International Conference on Machine Learning (ICML), PMLR 139, 8748–8763. link
  2. Ngiam, J., Khosla, A., Kim, M., Nam, J., Lee, H., & Ng, A. Y. (2011). Multimodal deep learning. Proceedings of the 28th International Conference on Machine Learning (ICML), 689–696. link

इस पृष्ठ का उद्धरण कैसे दें

ScholarGate. (2026, June 3). Multimodal Image Classification (Vision + Auxiliary Modality Fusion). ScholarGate. https://scholargate.app/hi/deep-learning/multimodal-image-classification

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

Compare side by side

इनमें संदर्भित

ScholarGateMultimodal Image Classification (Multimodal Image Classification (Vision + Auxiliary Modality Fusion)). 2026-06-15 को यहाँ से प्राप्त https://scholargate.app/hi/deep-learning/multimodal-image-classification · डेटासेट: https://doi.org/10.5281/zenodo.20539026