Official checkpoints for Trans4trans: Efficient Transformer for Transparent Object & Scene Segmentation (ICCVW 2021, T-ITS 2022).
AI & ML interests
None defined yet.
Recent Activity
Trans4Map: Revisiting Holistic Bird's-Eye-View Mapping from Egocentric Images to Allocentric Semantics with Vision Transformers (WACV 2023).
Event-based semantic segmentation for accident scenes under extreme dynamics and motion blur (IROS 2021 & T-ITS 2021).
Arbitrary-modal semantic segmentation suite (Depth, LiDAR, Events, Multi-Views, RGB) and CMNeXt models.
Official checkpoints for Trans4trans: Efficient Transformer for Transparent Object & Scene Segmentation (ICCVW 2021, T-ITS 2022).
MatchFormer: Interleaving Attention in Transformers for Feature Matching (ACCV 2022).
Trans4Map: Revisiting Holistic Bird's-Eye-View Mapping from Egocentric Images to Allocentric Semantics with Vision Transformers (WACV 2023).
Panoramic Semantic Mapping for Indoor Bird's-Eye View (IEEE/CVF WACV 2024)
Event-based semantic segmentation for accident scenes under extreme dynamics and motion blur (IROS 2021 & T-ITS 2021).
Panoramic semantic segmentation and domain adaptation suite with Trans4PASS, Trans4PASS+, and SynPASS benchmark.
Arbitrary-modal semantic segmentation suite (Depth, LiDAR, Events, Multi-Views, RGB) and CMNeXt models.