The rapid growth of the internet and technology has led to an exponential increase in the volume of information individuals must manage, which results in a rising demand for search engines to assess relevance quickly. Hence, information retrieval has garnered increasing attention within research communities. The Video Browser Showdown (VBS) is an annual competition that allows researchers to evaluate their methodologies against established benchmarks. In this work, we introduce IMSearch 2.0, an enhanced interactive video retrieval system for VBS 2025. This revised version features several improvements aimed at assisting users in effectively addressing multimedia retrieval tasks. Particularly, IMSearch 2.0 retains the core search function across multiple modalities, including text, automatic speech recognition (ASR) and optical character recognition (OCR), followed by a relevance feedback mechanism. Additionally, we develop a new semi-automatic keyframe selection, a colour-dominant search and replace the traditional CLIP embedding methods with a more effective model to achieve a more comprehensive visual representation. Finally, the user interface has been modified to be more user-centric and easy-to-use for all users.