Infrared and visible image fusion aims to integrate the complementary information from thermal radiation and reflectance imaging， while highlighting salient targets and preserving texture details， thereby providing more comprehensive inputs for human visual perception and downstream vision tasks. To this end， this paper proposes a segmentation- and detection-driven infrared and visible image fusion network. The framework consists of a fusion network and two task-driven branches： an object discriminator and a segmentation branch， which guide the fusion network to retain richer high-level semantics through their respective loss functions. To enhance feature representation， we design a Dense Connection and Gradient Residual Module （DCGRM） based on dense blocks for deep feature extraction. Furthermore， a Large Kernel Attention （LKA） module is introduced in the decoding stage to focus on key regions and reduce information loss， further improving the quality of the fused images. Experiments demonstrate that the proposed method not only highlights salient targets but also preserves rich details， outperforming comparative methods across multiple fusion metrics. Moreover， benefiting from the task-driven collaborative design， the proposed method also exhibits superior performance in downstream vision tasks， including segmentation and object detection.