-- We introduce DexGanGrasp, a dexterous grasp synthesis method that generates and evaluates grasps with a single view in real-time. DexGanGrasp comprises a Conditional Generative Adversarial Network (cGAN)-based DexGenerator to generate dexterous grasps and a discriminator-like DexE-valautor to assess the stability of these grasps. Extensive simulation and real-world experiments showcase the effectiveness of our proposed method, outperforming the baseline FFHNet with an 18 . T o further achieve task-oriented grasping, we extend Dex-GanGrasp to DexAfford-Prompt, an open-vocabulary affordance grounding pipeline for dexterous grasping leveraging Multi-modal Large Language Models (MLLM) and Vision Language Models (VLM) with successful real-world deployments. As a crucial prerequisite for manipulation, robotic grasping is a fundamental skill for robots to start interacting with our environment. Low-dimensional end effectors like 2-jaw grippers or suction cups have been extensively studied with significant progress [1]-[4]. However, 2-jaw grippers suffer from low dexterity, which limits their applicability for task-oriented manipulations [5], [6]. Dexterous hands allow for a broader range of solutions to grasp an object.

We introduce DexDiffuser, a novel dexterous grasping method that generates, evaluates, and refines grasps on partial object point clouds. DexDiffuser includes the conditional diffusion-based grasp sampler DexSampler and the dexterous grasp evaluator DexEvaluator. DexSampler generates high-quality grasps conditioned on object point clouds by iterative denoising of randomly sampled grasps. We also introduce two grasp refinement strategies: Evaluator-Guided Diffusion (EGD) and Evaluator-based Sampling Refinement (ESR). Our simulation and real-world experiments on the Allegro Hand consistently demonstrate that DexDiffuser outperforms the state-of-the-art multi-finger grasp generation method FFHNet with an, on average, 21.71--22.20\% higher grasp success rate.