Pengfei Wu

vision transformer (1) curriculum learning (1) self-supervised learning (1) knowledge distillation (1) question answering (1) dialogue generation (1) multimodal learning (1) video understanding (1) person re-identification (1) adversarial training (1) temporal grounding (1) occlusion handling (1) self-paced learning (1) multimodal large language model (1) temporal attention (1) feature aggregation (1) context window (1) length extrapolation (1) token prediction (1) inference acceleration (1)