Mina Alibeigi

Mina Alibeigi

Auto-generated profile, based on publicly available information

Mina Alibeigi

Zenseact AB

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

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

Representative Publications

View All Publications
9 September 2023 in ACM Transactions on Embedded Computing Systems
Verified
Journal Article
Open access
SMSeyedhamidreza Mousavi
Seyedhamidreza Mousavi
MLMohammad Loni
Mohammad Loni
MAMina Alibeigi
Mina Alibeigi
Masoud DaneshtalabMasoud Daneshtalab
Masoud Daneshtalab
The deployment of Deep Neural Networks (DNNs) on edge devices is hindered by the substantial gap between performance requirements and available computational power. While recent research has made significant strides in developing pruning methods to build a sparse network for reducing the computing overhead of DNNs, there remains considerable accuracy loss, especially at high pruning ratios. We find that the architectures designed for dense networks by differentiable architecture search methods are ineffective when pruning mechanisms are applied to them. The main reason is that the current methods do not support sparse architectures in their search space and use a search objective that is made for dense networks and does not focus on sparsity. This paper proposes a new method to search for sparsity-friendly neural architectures. It is done by adding two new sparse operations to the search space and modifying the search objective. We propose two novel parametric SparseConv and SparseLinear operations in order to expand the search space to include sparse operations. In particular, these operations make a flexible search space due to using sparse parametric versions of linear and convolution operations. The proposed search objective lets us train the architecture based on the sparsity of the search space operations. Quantitative analyses demonstrate that architectures found through DASS outperform those used in the state-of-the-art sparse networks on the CIFAR-10 and ImageNet datasets. In terms of performance and hardware effectiveness, DASS increases the accuracy of the sparse version of MobileNet-v2 from 73.44% to 81.35% (+7.91% improvement) with a 3.87× faster inference time.

1 February 2017 in IEEE Transactions on Robotics
Verified
Journal Article
Closed access
MAMina Alibeigi
Mina Alibeigi
MAMajid Nili Ahmadabadi
Majid Nili Ahmadabadi
BABabak Nadjar Araabi
Babak Nadjar Araabi
Social robots are becoming a companion in everyday life. To be well accepted by humans, they should efficiently understand meanings of their partners' motions and body language and respond accordingly. Learning concepts by imitation brings them this ability in a user-friendly way. This paper presents a fast and robust model for incremental learning of concepts by imitation (ILoCI). In ILoCI, observed multimodal spatiotemporal demonstrations are incrementally abstracted and generalized based on their perceptual and functional similarities during the imitation. Perceptually similar demonstrations are abstracted by a dynamic model of the mirror neuron system. The functional similarities of demonstrations are also learned through a limited number of interactions with the teacher. Incremental relearning of acquired concepts together through memory rehearsal enables the learner to gradually extract and utilize the common structural relations among demonstrations to expedite the learning process especially at the initial stages. Performance of ILoCI is assessed using a standard benchmark dataset and a human-robot interaction task in which a humanoid robot learns to abstract teacher's hand motions during imitation. Its performance is also evaluated on occluded observations that are probable in real environments. The results show efficiency of ILoCI in concept acquisition, recognition, prediction, and generation in addition to its robustness to occlusions and high variability in observations.

7 September 2022
Verified
Book Chapter
Closed access
PDF file
AZAli Zoljodi
Ali Zoljodi
MLMohammad Loni
Mohammad Loni
SASadegh Abadijou
Sadegh Abadijou
MAMina Alibeigi
Mina Alibeigi
Masoud DaneshtalabMasoud Daneshtalab
Masoud Daneshtalab

12 September 2023 in arXiv
Verified
Preprint
Open access
PDF file
SMSeyedhamidreza Mousavi
Seyedhamidreza Mousavi
MLMohammad Loni
Mohammad Loni
MAMina Alibeigi
Mina Alibeigi
Masoud DaneshtalabMasoud Daneshtalab
Masoud Daneshtalab
The deployment of Deep Neural Networks (DNNs) on edge devices is hindered by the substantial gap between performance requirements and available processing power. While recent research has made significant strides in developing pruning methods to build a sparse network for reducing the computing overhead of DNNs, there remains considerable accuracy loss, especially at high pruning ratios. We find that the architectures designed for dense networks by differentiable architecture search methods are ineffective when pruning mechanisms are applied to them. The main reason is that the current method does not support sparse architectures in their search space and uses a search objective that is made for dense networks and does not pay any attention to sparsity. In this paper, we propose a new method to search for sparsity-friendly neural architectures. We do this by adding two new sparse operations to the search space and modifying the search objective. We propose two novel parametric SparseConv and SparseLinear operations in order to expand the search space to include sparse operations. In particular, these operations make a flexible search space due to using sparse parametric versions of linear and convolution operations. The proposed search objective lets us train the architecture based on the sparsity of the search space operations. Quantitative analyses demonstrate that our search architectures outperform those used in the stateof-the-art sparse networks on the CIFAR-10 and ImageNet datasets. In terms of performance and hardware effectiveness, DASS increases the accuracy of the sparse version of MobileNet-v2 from 73.44% to 81.35% (+7.91% improvement) with 3.87x faster inference time.