Picture for Hyunjun Kim

Hyunjun Kim

CODE: Contrasting Self-generated Description to Combat Hallucination in Large Multi-modal Models

Add code
Jun 04, 2024
Viaarxiv icon

Robust Pedestrian Detection via Constructing Versatile Pedestrian Knowledge Bank

Apr 30, 2024
Viaarxiv icon

On the Consideration of AI Openness: Can Good Intent Be Abused?

Add code
Mar 11, 2024
Figure 1 for On the Consideration of AI Openness: Can Good Intent Be Abused?
Figure 2 for On the Consideration of AI Openness: Can Good Intent Be Abused?
Figure 3 for On the Consideration of AI Openness: Can Good Intent Be Abused?
Figure 4 for On the Consideration of AI Openness: Can Good Intent Be Abused?
Viaarxiv icon

Incorporating Language-Driven Appearance Knowledge Units with Visual Cues in Pedestrian Detection

Nov 02, 2023
Viaarxiv icon

Speaker-adaptive Lip Reading with User-dependent Padding

Aug 09, 2022
Figure 1 for Speaker-adaptive Lip Reading with User-dependent Padding
Figure 2 for Speaker-adaptive Lip Reading with User-dependent Padding
Figure 3 for Speaker-adaptive Lip Reading with User-dependent Padding
Figure 4 for Speaker-adaptive Lip Reading with User-dependent Padding
Viaarxiv icon

Fast Monte-Carlo Approximation of the Attention Mechanism

Jan 30, 2022
Figure 1 for Fast Monte-Carlo Approximation of the Attention Mechanism
Figure 2 for Fast Monte-Carlo Approximation of the Attention Mechanism
Figure 3 for Fast Monte-Carlo Approximation of the Attention Mechanism
Figure 4 for Fast Monte-Carlo Approximation of the Attention Mechanism
Viaarxiv icon