IEEE

Exploring Fine-Grained Image-Text Alignment for Referring Remote Sensing Image Segmentation

Abstract: Given a language expression, referring remote sensing image segmentation (RRSIS) aims to identify ground objects and assign pixelwise labels within the imagery. One of the key challenges for ...
IEEE

FGR-GA: Integrating Fine-Grained Representation Refinement and Group-Aware Alignment for Remote Sensing Image-Text Retrieval

Abstract: Leveraging powerful semantic understanding and generation capabilities, Vision-Language Pre-trained (VLP) large models have demonstrated remarkable potential in cross-modal retrieval.
GitHub

Visual-Text Cross Alignment: Refining the Similarity Score in Vision-Language Models (ICML2024)

Recent research shows that using a pre-trained vision-language model (VLM), like CLIP, to align a query image with detailed text descriptions generated by LLMs can enhance zero-shot classification.
GitHub

tutorial.py

img1 = cv2.imread('./Exp 04 - Matching and Alignment/porsche1.png') img2 = cv2.imread('./Exp 04 - Matching and Alignment/porsche2.png') img1 = cv2.resize(img1, (0, 0 ...