DETRs Beat YOLOs on Real-time Object Detection

Recently, end-to-end transformer-based detectors (DETRs) have achieved remarkable performance. However, the issue of the high computational cost of DETRs has not been effectively addressed, limiting their practical application and preventing them from fully exploiting the benefits of no post-processing, such as non-maximum suppression (NMS). In this paper, we first analyze the influence of NMS in modern real-time object detectors on inference speed, and establish an end-to-end speed benchmark. To avoid the inference delay caused by NMS, we propose a Real-Time DEtection TRansformer (RT-DETR), the first real-time end-to-end object detector to our best knowledge. Specifically, we design an efficient hybrid encoder to efficiently process multi-scale features by decoupling the intra-scale interaction and cross-scale fusion, and propose IoU-aware query selection to improve the initialization of object queries. In addition, our proposed detector supports flexibly adjustment of the inference speed by using different decoder layers without the need for retraining, which facilitates the practical application of real-time object detectors. Our RT-DETR-L achieves 53.0% AP on COCO val2017 and 114 FPS on T4 GPU, while RT-DETR-X achieves 54.8% AP and 74 FPS, outperforming all YOLO detectors of the same scale in both speed and accuracy. Furthermore, our RT-DETR-R50 achieves 53.1% AP and 108 FPS, outperforming DINO-Deformable-DETR-R50 by 2.2% AP in accuracy and by about 21 times in FPS. Source code and pretrained models will be available at PaddleDetection.

Multi-granularity Interaction Simulation for Unsupervised Interactive Segmentation

Interactive segmentation enables users to segment as needed by providing cues of objects, which introduces human-computer interaction for many fields, such as image editing and medical image analysis. Typically, massive and expansive pixel-level annotations are spent to train deep models by object-oriented interactions with manually labeled object masks. In this work, we reveal that informative interactions can be made by simulation with semantic-consistent yet diverse region exploration in an unsupervised paradigm. Concretely, we introduce a Multi-granularity Interaction Simulation (MIS) approach to open up a promising direction for unsupervised interactive segmentation. Drawing on the high-quality dense features produced by recent self-supervised models, we propose to gradually merge patches or regions with similar features to form more extensive regions and thus, every merged region serves as a semantic-meaningful multi-granularity proposal. By randomly sampling these proposals and simulating possible interactions based on them, we provide meaningful interaction at multiple granularities to teach the model to understand interactions. Our MIS significantly outperforms non-deep learning unsupervised methods and is even comparable with some previous deep-supervised methods without any annotation.

Position Embedding Needs an Independent Layer Normalization

The Position Embedding (PE) is critical for Vision Transformers (VTs) due to the permutation-invariance of self-attention operation. By analyzing the input and output of each encoder layer in VTs using reparameterization and visualization, we find that the default PE joining method (simply adding the PE and patch embedding together) operates the same affine transformation to token embedding and PE, which limits the expressiveness of PE and hence constrains the performance of VTs. To overcome this limitation, we propose a simple, effective, and robust method. Specifically, we provide two independent layer normalizations for token embeddings and PE for each layer, and add them together as the input of each layer's Muti-Head Self-Attention module. Since the method allows the model to adaptively adjust the information of PE for different layers, we name it as Layer-adaptive Position Embedding, abbreviated as LaPE. Extensive experiments demonstrate that LaPE can improve various VTs with different types of PE and make VTs robust to PE types. For example, LaPE improves 0.94% accuracy for ViT-Lite on Cifar10, 0.98% for CCT on Cifar100, and 1.72% for DeiT on ImageNet-1K, which is remarkable considering the negligible extra parameters, memory and computational cost brought by LaPE. The code is publicly available at https://github.com/Ingrid725/LaPE.

Dynamic Clustering Network for Unsupervised Semantic Segmentation

Recently, the ability of self-supervised Vision Transformer (ViT) to represent pixel-level semantic relationships promotes the development of unsupervised dense prediction tasks. In this work, we investigate transferring self-supervised ViT to unsupervised semantic segmentation task. According to the analysis that the pixel-level representations of self-supervised ViT within a single image achieve good intra-class compactness and inter-class discrimination, we propose the Dynamic Clustering Network (DCN) to dynamically infer the underlying cluster centers for different images. By training with the proposed modularity loss, the DCN learns to project a set of prototypes to cluster centers for pixel representations in each image and assign pixels to different clusters, resulting on dividing each image to class-agnostic regions. For achieving unsupervised semantic segmentation task, we treat it as a region classification problem. Based on the regions produced by the DCN, we explore different ways to extract region-level representations and classify them in an unsupervised manner. We demonstrate the effectiveness of the proposed method trough experiments on unsupervised semantic segmentation, and achieve state-of-the-art performance on PASCAL VOC 2012 unsupervised semantic segmentation task.

Difference in Euclidean Norm Can Cause Semantic Divergence in Batch Normalization

In this paper, we show that the difference in Euclidean norm of samples can make a contribution to the semantic divergence and even confusion, after the spatial translation and scaling transformation in batch normalization. To address this issue, we propose an intuitive but effective method to equalize the Euclidean norms of sample vectors. Concretely, we $l_2$-normalize each sample vector before batch normalization, and therefore the sample vectors are of the same magnitude. Since the proposed method combines the $l_2$ normalization and batch normalization, we name our method as $L_2$BN. The $L_2$BN can strengthen the compactness of intra-class features and enlarge the discrepancy of inter-class features. In addition, it can help the gradient converge to a stable scale. The $L_2$BN is easy to implement and can exert its effect without any additional parameters and hyper-parameters. Therefore, it can be used as a basic normalization method for neural networks. We evaluate the effectiveness of $L_2$BN through extensive experiments with various models on image classification and acoustic scene classification tasks. The experimental results demonstrate that the $L_2$BN is able to boost the generalization ability of various neural network models and achieve considerable performance improvements.