2023-03-20 更新
Commonsense Knowledge Assisted Deep Learning for Resource-constrained and Fine-grained Object Detection
Authors:Pu Zhang, Bin Liu
In this paper, we consider fine-grained image object detection in resource-constrained cases such as edge computing. Deep learning (DL), namely learning with deep neural networks (DNNs), has become the dominating approach to object detection. To achieve accurate fine-grained detection, one needs to employ a large enough DNN model and a vast amount of data annotations, which brings a challenge for using modern DL object detectors in resource-constrained cases. To this end, we propose an approach, which leverages commonsense knowledge to assist a coarse-grained object detector to get accurate fine-grained detection results. Specifically, we introduce a commonsense knowledge inference module (CKIM) to process coarse-grained lables given by a benchmark DL detector to produce fine-grained lables. We consider both crisp-rule and fuzzy-rule based inference in our CKIM; the latter is used to handle ambiguity in the target semantic labels. We implement our method based on several modern DL detectors, namely YOLOv4, Mobilenetv3-SSD and YOLOv7-tiny. Experiment results show that our approach outperforms benchmark detectors remarkably in terms of accuracy, model size and processing latency.
PDF 7 pages
点此查看论文截图
Scribble-Supervised RGB-T Salient Object Detection
Authors:Zhengyi Liu, Xiaoshen Huang, Guanghui Zhang, Xianyong Fang, Linbo Wang, Bin Tang
Salient object detection segments attractive objects in scenes. RGB and thermal modalities provide complementary information and scribble annotations alleviate large amounts of human labor. Based on the above facts, we propose a scribble-supervised RGB-T salient object detection model. By a four-step solution (expansion, prediction, aggregation, and supervision), label-sparse challenge of scribble-supervised method is solved. To expand scribble annotations, we collect the superpixels that foreground scribbles pass through in RGB and thermal images, respectively. The expanded multi-modal labels provide the coarse object boundary. To further polish the expanded labels, we propose a prediction module to alleviate the sharpness of boundary. To play the complementary roles of two modalities, we combine the two into aggregated pseudo labels. Supervised by scribble annotations and pseudo labels, our model achieves the state-of-the-art performance on the relabeled RGBT-S dataset. Furthermore, the model is applied to RGB-D and video scribble-supervised applications, achieving consistently excellent performance.
PDF ICME2023
点此查看论文截图
Exploring Sparse Visual Prompt for Cross-domain Semantic Segmentation
Authors:Senqiao Yang, Jiarui Wu, Jiaming Liu, Xiaoqi Li, Qizhe Zhang, Mingjie Pan, Mingjie Pan, Shanghang Zhang
Visual Domain Prompts (VDP) have shown promising potential in addressing visual cross-domain problems. Existing methods adopt VDP in classification domain adaptation (DA), such as tuning image-level or feature-level prompts for target domains. Since the previous dense prompts are opaque and mask out continuous spatial details in the prompt regions, it will suffer from inaccurate contextual information extraction and insufficient domain-specific feature transferring when dealing with the dense prediction (i.e. semantic segmentation) DA problems. Therefore, we propose a novel Sparse Visual Domain Prompts (SVDP) approach tailored for addressing domain shift problems in semantic segmentation, which holds minimal discrete trainable parameters (e.g. 10\%) of the prompt and reserves more spatial information. To better apply SVDP, we propose Domain Prompt Placement (DPP) method to adaptively distribute several SVDP on regions with large data distribution distance based on uncertainty guidance. It aims to extract more local domain-specific knowledge and realizes efficient cross-domain learning. Furthermore, we design a Domain Prompt Updating (DPU) method to optimize prompt parameters differently for each target domain sample with different degrees of domain shift, which helps SVDP to better fit target domain knowledge. Experiments, which are conducted on the widely-used benchmarks (Cityscapes, Foggy-Cityscapes, and ACDC), show that our proposed method achieves state-of-the-art performances on the source-free adaptations, including six Test Time Adaptation and one Continual Test-Time Adaptation in semantic segmentation.
PDF
点此查看论文截图
Revisiting Image Reconstruction for Semi-supervised Semantic Segmentation
Authors:Yuhao Lin, Haiming Xu, Lingqiao Liu, Jinan Zou, Javen Qinfeng Shi
Autoencoding, which aims to reconstruct the input images through a bottleneck latent representation, is one of the classic feature representation learning strategies. It has been shown effective as an auxiliary task for semi-supervised learning but has become less popular as more sophisticated methods have been proposed in recent years. In this paper, we revisit the idea of using image reconstruction as the auxiliary task and incorporate it with a modern semi-supervised semantic segmentation framework. Surprisingly, we discover that such an old idea in semi-supervised learning can produce results competitive with state-of-the-art semantic segmentation algorithms. By visualizing the intermediate layer activations of the image reconstruction module, we show that the feature map channel could correlate well with the semantic concept, which explains why joint training with the reconstruction task is helpful for the segmentation task. Motivated by our observation, we further proposed a modification to the image reconstruction task, aiming to further disentangle the object clue from the background patterns. From experiment evaluation on various datasets, we show that using reconstruction as auxiliary loss can lead to consistent improvements in various datasets and methods. The proposed method can further lead to significant improvement in object-centric segmentation tasks.
PDF
点此查看论文截图
Enhancing the Role of Context in Region-Word Alignment for Object Detection
Authors:Kyle Buettner, Adriana Kovashka
Vision-language pretraining to learn a fine-grained, region-word alignment between image-caption pairs has propelled progress in open-vocabulary object detection. We observe that region-word alignment methods are typically used in detection with respect to only object nouns, and the impact of other rich context in captions, such as attributes, is unclear. In this study, we explore how language context affects downstream object detection and propose to enhance the role of context. In particular, we show how to strategically contextualize the grounding pretraining objective for improved alignment. We further hone in on attributes as especially useful object context and propose a novel adjective and noun-based negative sampling strategy for increasing their focus in contrastive learning. Overall, our methods enhance object detection when compared to the state-of-the-art in region-word pretraining. We also highlight the fine-grained utility of an attribute-sensitive model through text-region retrieval and phrase grounding analysis.
PDF
点此查看论文截图