DeoT: an end-to-end encoder-only Transformer object detector

Ding, Tonghe; Feng, Kaili; Wei, Yanjun; Han, Yonghao; Li, Tianping

doi:10.1007/s11554-023-01280-0

Cited by 3 publications

(2 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Data augmentation can be seen as an indirect approach to minimize the gap between train and test sets [83]. Several methods used augmentation in their detection task including T-TRD [43], SPH-Yolov5 [73], MATR [71], NLFFTNet [84], DeoT [85], HTDet [86], and Sw-YoloX [63]. (vi) One-to-Many Label Assignment: The one-to-one matching in DETR can result in poor discriminative features within the encoder.…”

Section: Auxiliary Techniquesmentioning

confidence: 99%

“…NLFFTNet [84] addresses the limitation of only considering local interactions in current fusion techniques by introducing a nonlocal feature-fused transformer convolutional network, capturing longdistance semantic relationships between different feature layers. DeoT [85] merges an encoder-only transformer with a novel feature pyramid fusion module. This fusion is enhanced by the use of channel and spatial attention in the Channel Refinement Module (CRM) and Spatial Refinement Module (SRM), enabling the extraction of richer features.…”

Section: Improved Feature Representationmentioning

confidence: 99%

See 1 more Smart Citation

Extended Expectation Maximization for Under-Fitted Models

Rekavandi¹,

SEGHOUANE²,

Boussaid³

et al. 2022

Preprint

View full text Add to dashboard Cite

<p>In this paper, we generalize the well-known Expectation Maximization (EM) algorithm using the α− divergence for Gaussian Mixture Model (GMM). This approach is used in robust subspace detection when the number of parameters is kept small to avoid overfitting and large estimation variances. The level of robustness can be tuned by the parameterα. When α→1, our method is equivalent to the standard EM approach and for α <1 the method is robust against potential outliers. Simulation results show that the method outperforms the standard EM when it comes to mismatches between noise models and their realizations. In addition, we use the proposed method to detect active brain areas using collected functional Magnetic Resonance Imaging (fMRI) data during task-related experiments.</p>

show abstract