Learning imbalanced data with vision transformers
The real-world data tends to be heavily imbalanced and severely skew the data-driven deep
neural networks, which makes Long-Tailed Recognition (LTR) a massive challenging task …
neural networks, which makes Long-Tailed Recognition (LTR) a massive challenging task …
Ti-mae: Self-supervised masked time series autoencoders
Multivariate Time Series forecasting has been an increasingly popular topic in various
applications and scenarios. Recently, contrastive learning and Transformer-based models …
applications and scenarios. Recently, contrastive learning and Transformer-based models …
Curriculum learning with infant egocentric videos
S Sheybani, H Hansaria, J Wood… - Advances in Neural …, 2024 - proceedings.neurips.cc
Infants possess a remarkable ability to rapidly learn and process visual inputs. As an infant's
mobility increases, so does the variety and dynamics of their visual inputs. Is this change in …
mobility increases, so does the variety and dynamics of their visual inputs. Is this change in …
Interacting-enhancing feature transformer for cross-modal remote-sensing image and text retrieval
Cross-modal remote-sensing image–text retrieval (CMRSITR) is a challenging topic in the
remote-sensing (RS) community. It has gained growing attention because it can be flexibly …
remote-sensing (RS) community. It has gained growing attention because it can be flexibly …
1% vs 100%: Parameter-efficient low rank adapter for dense predictions
Fine-tuning large-scale pre-trained vision models to downstream tasks is a standard
technique for achieving state-of-the-art performance on computer vision benchmarks …
technique for achieving state-of-the-art performance on computer vision benchmarks …
Timebalance: Temporally-invariant and temporally-distinctive video representations for semi-supervised action recognition
Abstract Semi-Supervised Learning can be more beneficial for the video domain compared
to images because of its higher annotation cost and dimensionality. Besides, any video …
to images because of its higher annotation cost and dimensionality. Besides, any video …
How can objects help action recognition?
Current state-of-the-art video models process a video clip as a long sequence of spatio-
temporal tokens. However, they do not explicitly model objects, their interactions across the …
temporal tokens. However, they do not explicitly model objects, their interactions across the …
Multi-Space Alignments Towards Universal LiDAR Segmentation
A unified and versatile LiDAR segmentation model with strong robustness and
generalizability is desirable for safe autonomous driving perception. This work presents …
generalizability is desirable for safe autonomous driving perception. This work presents …
Does visual pretraining help end-to-end reasoning?
We aim to investigate whether end-to-end learning of visual reasoning can be achieved with
general-purpose neural networks, with the help of visual pretraining. A positive result would …
general-purpose neural networks, with the help of visual pretraining. A positive result would …
Asymmetric masked distillation for pre-training small foundation models
Self-supervised foundation models have shown great potential in computer vision thanks to
the pre-training paradigm of masked autoencoding. Scale is a primary factor influencing the …
the pre-training paradigm of masked autoencoding. Scale is a primary factor influencing the …