Picture for Peng Li

Peng Li

DJI Innovations Inc

Perspective Transition of Large Language Models for Solving Subjective Tasks

Add code
Jan 16, 2025
Viaarxiv icon

Hierarchical Superpixel Segmentation via Structural Information Theory

Add code
Jan 13, 2025
Viaarxiv icon

Diffusion as Shader: 3D-aware Video Diffusion for Versatile Video Generation Control

Add code
Jan 07, 2025
Viaarxiv icon

Rate-Distortion Optimized Skip Coding of Region Adaptive Hierarchical Transform Coefficients for MPEG G-PCC

Add code
Dec 07, 2024
Viaarxiv icon

Towards 3D Acceleration for low-power Mixture-of-Experts and Multi-Head Attention Spiking Transformers

Add code
Dec 07, 2024
Viaarxiv icon

Trimming Down Large Spiking Vision Transformers via Heterogeneous Quantization Search

Add code
Dec 07, 2024
Figure 1 for Trimming Down Large Spiking Vision Transformers via Heterogeneous Quantization Search
Figure 2 for Trimming Down Large Spiking Vision Transformers via Heterogeneous Quantization Search
Figure 3 for Trimming Down Large Spiking Vision Transformers via Heterogeneous Quantization Search
Figure 4 for Trimming Down Large Spiking Vision Transformers via Heterogeneous Quantization Search
Viaarxiv icon

Align3R: Aligned Monocular Depth Estimation for Dynamic Videos

Add code
Dec 05, 2024
Viaarxiv icon

Interactive Visual Assessment for Text-to-Image Generation Models

Add code
Nov 23, 2024
Viaarxiv icon

AIGS: Generating Science from AI-Powered Automated Falsification

Add code
Nov 17, 2024
Figure 1 for AIGS: Generating Science from AI-Powered Automated Falsification
Figure 2 for AIGS: Generating Science from AI-Powered Automated Falsification
Figure 3 for AIGS: Generating Science from AI-Powered Automated Falsification
Figure 4 for AIGS: Generating Science from AI-Powered Automated Falsification
Viaarxiv icon

Spiking Transformer Hardware Accelerators in 3D Integration

Add code
Nov 11, 2024
Viaarxiv icon