Object detection using RGB images has achieved significant progress, driven by the availability of extensive data and established models. Despite these advancements, object detection across alternative imaging modalities such as thermal/infrared, ultraviolet, X-ray, MRI, and acoustic imaging continues to lag, largely due to the lack of labeled datasets. This paper presents a simple method that leverages intermediate feature maps from RGB object detection to improve performance in thermal imaging, where data constraints are prevalent. Our approach seeks to narrow the research divide between the well-established RGB detection and the relatively nascent field of thermal object detection. We validate the effectiveness of our method using the FLIR dataset, and we will provide the source code at: https://github.com/redacted for anonymity.