Mengke Li

Mengke Li

Auto-generated profile, based on publicly available information

Mengke Li

College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, China

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 April 2024 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Open access
ZHZhikai Hu
Zhikai Hu
YCYiu-ming Cheung
Yiu-ming Cheung
MLMengke Li
Mengke Li
WLWeichao Lan
Weichao Lan
Cross-modal hashing (CMH) has attracted considerable attention in recent years. Almost all existing CMH methods primarily focus on reducing the modality gap and semantic gap, i.e., aligning multi-modal features and their semantics in Hamming space, without taking into account the space gap, i.e., difference between the real number space and the Hamming space. In fact, the space gap can affect the performance of CMH methods. In this paper, we analyze and demonstrate how the space gap affects the existing CMH methods, which therefore raises two problems: solution space compression and loss function oscillation. These two problems eventually cause the retrieval performance deteriorating. Based on these findings, we propose a novel algorithm, namely Semantic Channel Hashing (SCH). First, we classify sample pairs into fully semantic-similar, partially semantic-similar, and semantic-negative ones based on their similarity and impose different constraints on them, respectively, to ensure that the entire Hamming space is utilized. Then, we introduce a semantic channel to alleviate the issue of loss function oscillation. Experimental results on three public datasets demonstrate that SCH outperforms the state-of-the-art methods. Furthermore, experimental validations are provided to substantiate the conjectures regarding solution space compression and loss function oscillation, offering visual evidence of their impact on the CMH methods.

3 August 2022 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
MLMengke Li
Mengke Li
YCYiu-ming Cheung
Yiu-ming Cheung
ZHZhikai Hu
Zhikai Hu
For long-tailed distributed data, existing classification models often learn overwhelmingly on the head classes while ignoring the tail classes, resulting in poor generalization capability. To address this problem, we thereby propose a new approach in this paper, in which a key point sensitive (KPS) loss is presented to regularize the key points strongly to improve the generalization performance of the classification model. Meanwhile, in order to improve the performance on tail classes, the proposed KPS loss also assigns relatively large margins on tail classes. Furthermore, we propose a gradient adjustment (GA) optimization strategy to re-balance the gradients of positive and negative samples for each class. By virtue of the gradient analysis of the loss function, it is found that the tail classes always receive negative signals during training, which misleads the tail prediction to be biased towards the head. The proposed GA strategy can circumvent excessive negative signals on tail classes and further improve the overall classification accuracy. Extensive experiments conducted on long-tailed benchmarks show that the proposed method is capable of significantly improving the classification accuracy of the model in tail classes while maintaining competent performance in head classes.

11 April 2025 in IEEE Transactions on Cognitive and Developmental Systems
Verified
Journal Article
Open access
WLWeichao Lan
Weichao Lan
YCYiu-ming Cheung
Yiu-ming Cheung
QXQing Xu
Qing Xu
BLBuhua Liu
Buhua Liu
ZHZhikai Hu
Zhikai Hu
Zhenghua ChenZhenghua Chen
Zhenghua Chen
Knowledge distillation (KD) transferring knowledge from a large teacher model to a lightweight student one has received great attention in deep model compression. In addition to the supervision of ground truth, the vanilla KD method regards the predictions of the teacher as soft labels to supervise the training of the student model. Based on vanilla KD, various approaches have been developed to further improve the performance of the student model. However, few of these previous methods have considered the reliability of the supervision from teacher models. Supervision from erroneous predictions may mislead the training of the student model. This paper therefore proposes to tackle this problem from two aspects: Label Revision to rectify the incorrect supervision and Data Selection to select appropriate samples for distillation to reduce the impact of erroneous supervision. In the former, we propose to rectify the teacher’s inaccurate predictions using the ground truth. In the latter, we introduce a data selection technique to choose suitable training samples to be supervised by the teacher, thereby reducing the impact of incorrect predictions to some extent. Experiment results demonstrate the effectiveness of the proposed method, which can be further combined with other distillation approaches to enhance their performance.

15 May 2024 in IEEE Transactions on Artificial Intelligence
Verified
Journal Article
Open access
MLMengke Li
Mengke Li
YCYiu-ming Cheung
Yiu-ming Cheung
Yang LuYang Lu
Yang Lu
ZHZhikai Hu
Zhikai Hu
WLWeichao Lan
Weichao Lan
HHHui Huang
Hui Huang
It is not uncommon that real-world data are distributed with a long tail. For such data, the learning of deep neural networks becomes challenging because it is hard to classify tail classes correctly. In the literature, several existing methods have addressed this problem by reducing classifier bias, provided that the features obtained with long-tailed data are representative enough. However, we find that training directly on long-tailed data leads to uneven embedding space. That is, the embedding space of head classes severely compresses that of tail classes, which is not conducive to subsequent classifier learning. This paper therefore studies the problem of long-tailed visual recognition from the perspective of feature level. We introduce feature augmentation to balance the embedding distribution. The features of different classes are perturbed with varying amplitudes in Gaussian form. Based on these perturbed features, two novel logit adjustment methods are proposed to improve model performance at a modest computational overhead. Subsequently, the distorted embedding spaces of all classes can be calibrated. In such balanced-distributed embedding spaces, the biased classifier can be eliminated by simply retraining the classifier with class-balanced sampling data. Extensive experiments conducted on benchmark datasets demonstrate the superior performance of the proposed method over the state-of-the-art ones. Source code is available at https://github.com/Keke921/GCLLoss .

17 August 2021 in IEEE Transactions on Emerging Topics in Computational Intelligence
Verified
Journal Article
Closed access
MLMengke Li
Mengke Li
YCYiu-ming Cheung
Yiu-ming Cheung
Complete face recovering (CFR) is to recover the face image of a given partial face image of a target person whose photo may not be included in the gallery set. CFR has several attractive potential applications in surveillance, personal identification in forensics, to name a few, but it is challenging because of little information revealed from a single partial face image. Furthermore, the facial identity may get lost when recovering the complete face image. As far as we know, CFR problem has yet to be explored in the literature. This paper therefore proposes an identity-preserved CFR approach (IP-CFR) to tackle this problem. Accordingly, a denoising auto-encoder based network is applied. We propose an identity-preserved loss function to constrain the features in latent space of decoder, whereby maintaining the personal identity information. Then, to better restore the complete face image, the acquired features are further fed into a decoder with an adversarial structure that takes a new variant of discriminator. That is, we borrow the idea from energy based GAN that utilize an auto-encoder structure discriminator. It can produce very different gradient directions within the minibatch and therefore can make the model be trained stably. Further, we propose a novel dual-pipeline structure in the discriminator, which is leveraged to enhance the quality of the recovered image. Experimental results on the benchmark datasets show the superiority of IP-CFR.

10 October 2023 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Open access
WLWeichao Lan
Weichao Lan
YCYiu-ming Cheung
Yiu-ming Cheung
JJJuyong Jiang
Juyong Jiang
ZHZhikai Hu
Zhikai Hu
MLMengke Li
Mengke Li
As an effective tool for network compression, pruning techniques have been widely used to reduce the large number of parameters in deep neural networks (NNs). Nevertheless, unstructured pruning has the limitation of dealing with the sparse and irregular weights. By contrast, structured pruning can help eliminate this drawback but it requires complex criteria to determine which components to be pruned. Therefore, this paper presents a new method termed BUnit-Net, which directly constructs compact NNs by stacking designed basic units, without requiring additional judgement criteria anymore. Given the basic units of various architectures, they are combined and stacked systematically to build up compact NNs which involve fewer weight parameters due to the independence among the units. In this way, BUnit-Net can achieve the same compression effect as unstructured pruning while the weight tensors can still remain regular and dense. We formulate BUnit-Net in diverse popular backbones in comparison with the state-of-the-art pruning methods on different benchmark datasets. Moreover, two new metrics are proposed to evaluate the trade-off of compression performance. Experiment results show that BUnit-Net can achieve comparable classification accuracy while saving around 80% FLOPs and 73% parameters. That is, stacking basic units provides a new promising way for network compression.