Qi Tian

22

Papers

2,064

Total Citations

Papers (22)

4D Gaussian Splatting for Real-Time Dynamic Scene Rendering

ControlVideo: Training-free Controllable Text-to-video Generation

GaussianDreamer: Fast Generation from Text to 3D Gaussians by Bridging 2D and 3D Diffusion Models

GaussianEditor: Editing 3D Gaussians Delicately with Text Instructions

Towards 3D Molecule-Text Interpretation in Language Models

LogFormer: A Pre-train and Tuning Pipeline for Log Anomaly Detection

Improving Image Restoration through Removing Degradations in Textual Representations

Enhance Image Classification via Inter-Class Image Mixup with Diffusion Model

LION: Implicit Vision Prompt Tuning

C-CLIP: Multimodal Continual Learning for Vision-Language Model

Hybrid Distillation: Connecting Masked Autoencoders with Contrastive Learners

AlignZeg: Mitigating Objective Misalignment for Zero-shot Semantic Segmentation

Boosting Segment Anything Model Towards Open-Vocabulary Learning

METEOR: Multi-Encoder Collaborative Token Pruning for Efficient Vision Language Models

Optimize Incompatible Parameters Through Compatibility-aware Knowledge Integration

Parameter Efficient Fine-tuning via Cross Block Orchestration for Segment Anything Model

Infinite-Canvas: Higher-Resolution Video Outpainting with Extensive Content Generation

Segment Any 3D Gaussians

IM-Zero: Instance-level Motion Controllable Video Generation in a Zero-shot Manner

CLIP-Adapted Region-to-Text Learning for Generative Open-Vocabulary Semantic Segmentation

HalluciDoctor: Mitigating Hallucinatory Toxicity in Visual Instruction Data

OVMR: Open-Vocabulary Recognition with Multi-Modal References