Hamed R. Tavakoli

Andrey Moskalenko, Alexey Bryncev, Dmitry Vatolin et al.

This paper reviews the Challenge on Video Saliency Prediction at AIM 2024. The goal of the participants was to develop a method for predicting accurate saliency maps for the provided set of video sequences. Saliency maps are widely exploited in various applications, including video compression, quality assessment, visual perception studies, the advertising industry, etc. For this competition, a previously unused large-scale audio-visual mouse saliency (AViMoS) dataset of 1500 videos with more than 70 observers per video was collected using crowdsourced mouse tracking. The dataset collection methodology has been validated using conventional eye-tracking data and has shown high consistency. Over 30 teams registered in the challenge, and there are 7 teams that submitted the results in the final phase. The final phase solutions were tested and ranked by commonly used quality metrics on a private test subset. The results of this evaluation and the descriptions of the solutions are presented in this report. All data, including the private test subset, is made publicly available on the challenge homepage - https://challenges.videoprocessing.ai/challenges/video-saliency-prediction.html.

Juuso Korhonen, Goutham Rangu, Hamed R. Tavakoli et al.

We propose an application of online hard sample mining for efficient training of Neural Radiance Fields (NeRF). NeRF models produce state-of-the-art quality for many 3D reconstruction and rendering tasks but require substantial computational resources. The encoding of the scene information within the NeRF network parameters necessitates stochastic sampling. We observe that during the training, a major part of the compute time and memory usage is spent on processing already learnt samples, which no longer affect the model update significantly. We identify the backward pass on the stochastic samples as the computational bottleneck during the optimization. We thus perform the first forward pass in inference mode as a relatively low-cost search for hard samples. This is followed by building the computational graph and updating the NeRF network parameters using only the hard samples. To demonstrate the effectiveness of the proposed approach, we apply our method to Instant-NGP, resulting in significant improvements of the view-synthesis quality over the baseline (1 dB improvement on average per training time, or 2x speedup to reach the same PSNR level) along with approx. 40% memory savings coming from using only the hard samples to build the computational graph. As our method only interfaces with the network module, we expect it to be widely applicable.

Hamed R. Tavakoli, Ali Borji, Esa Rahtu et al.

This paper studies audio-visual deep saliency prediction. It introduces a conceptually simple and effective Deep Audio-Visual Embedding for dynamic saliency prediction dubbed ``DAVE" in conjunction with our efforts towards building an Audio-Visual Eye-tracking corpus named ``AVE". Despite existing a strong relation between auditory and visual cues for guiding gaze during perception, video saliency models only consider visual cues and neglect the auditory information that is ubiquitous in dynamic scenes. Here, we investigate the applicability of audio cues in conjunction with visual ones in predicting saliency maps using deep neural networks. To this end, the proposed model is intentionally designed to be simple. Two baseline models are developed on the same architecture which consists of an encoder-decoder. The encoder projects the input into a feature space followed by a decoder that infers saliency. We conduct an extensive analysis on different modalities and various aspects of multi-model dynamic saliency prediction. Our results suggest that (1) audio is a strong contributing cue for saliency prediction, (2) salient visible sound-source is the natural cause of the superiority of our Audio-Visual model, (3) richer feature representations for the input space leads to more powerful predictions even in absence of more sophisticated saliency decoders, and (4) Audio-Visual model improves over 53.54\% of the frames predicted by the best Visual model (our baseline). Our endeavour demonstrates that audio is an important cue that boosts dynamic video saliency prediction and helps models to approach human performance. The code is available at https://github.com/hrtavakoli/DAVE

Marco Filetti, Hamed R. Tavakoli, Niklas Ravaja et al.

PeyeDF is a Portable Document Format (PDF) reader with eye tracking support, available as free and open source software. It is especially useful to researchers investigating reading and learning phenomena, as it integrates PDF reading-related behavioural data with gaze-related data. It is suitable for short and long-term research and supports multiple eye tracking systems. We utilised it to conduct an experiment which demonstrated that features obtained from both gaze and reading data collected in the past can predict reading comprehension which takes place in the future. PeyeDF also provides an integrated means for data collection and indexing using the DiMe personal data storage system. It is designed to collect data in the background without interfering with the reading experience, behaving like a modern lightweight PDF reader. Moreover, it supports annotations, tagging and collaborative work. A modular design allows the application to be easily modified in order to support additional eye tracking protocols and run controlled experiments. We discuss the implementation of the software and report on the results of the experiment which we conducted with it.

Sen He, Hamed R. Tavakoli, Ali Borji et al.

In this work, we present a novel dataset consisting of eye movements and verbal descriptions recorded synchronously over images. Using this data, we study the differences in human attention during free-viewing and image captioning tasks. We look into the relationship between human attention and language constructs during perception and sentence articulation. We also analyse attention deployment mechanisms in the top-down soft attention approach that is argued to mimic human attention in captioning tasks, and investigate whether visual saliency can help image captioning. Our study reveals that (1) human attention behaviour differs in free-viewing and image description tasks. Humans tend to fixate on a greater variety of regions under the latter task, (2) there is a strong relationship between described objects and attended objects ($97\%$ of the described objects are being attended), (3) a convolutional neural network as feature encoder accounts for human-attended regions during image captioning to a great extent (around $78\%$), (4) soft-attention mechanism differs from human attention, both spatially and temporally, and there is low correlation between caption scores and attention consistency scores. These indicate a large gap between humans and machines in regards to top-down attention, and (5) by integrating the soft attention model with image saliency, we can significantly improve the model's performance on Flickr30k and MSCOCO benchmarks. The dataset can be found at: https://github.com/SenHe/Human-Attention-in-Image-Captioning.

Nannan Zou, Honglei Zhang, Francesco Cricri et al.

Neural image coding represents now the state-of-the-art image compression approach. However, a lot of work is still to be done in the video domain. In this work, we propose an end-to-end learned video codec that introduces several architectural novelties as well as training novelties, revolving around the concepts of adaptation and attention. Our codec is organized as an intra-frame codec paired with an inter-frame codec. As one architectural novelty, we propose to train the inter-frame codec model to adapt the motion estimation process based on the resolution of the input video. A second architectural novelty is a new neural block that combines concepts from split-attention based neural networks and from DenseNets. Finally, we propose to overfit a set of decoder-side multiplicative parameters at inference time. Through ablation studies and comparisons to prior art, we show the benefits of our proposed techniques in terms of coding gains. We compare our codec to VVC/H.266 and RLVC, which represent the state-of-the-art traditional and end-to-end learned codecs, respectively, and to the top performing end-to-end learned approach in 2021 CLIC competition, E2E_T_OL. Our codec clearly outperforms E2E_T_OL, and compare favorably to VVC and RLVC in some settings.

Honglei Zhang, Francesco Cricri, Hamed R. Tavakoli et al.

Lossless image compression is an important technique for image storage and transmission when information loss is not allowed. With the fast development of deep learning techniques, deep neural networks have been used in this field to achieve a higher compression rate. Methods based on pixel-wise autoregressive statistical models have shown good performance. However, the sequential processing way prevents these methods to be used in practice. Recently, multi-scale autoregressive models have been proposed to address this limitation. Multi-scale approaches can use parallel computing systems efficiently and build practical systems. Nevertheless, these approaches sacrifice compression performance in exchange for speed. In this paper, we propose a multi-scale progressive statistical model that takes advantage of the pixel-wise approach and the multi-scale approach. We developed a flexible mechanism where the processing order of the pixels can be adjusted easily. Our proposed method outperforms the state-of-the-art lossless image compression methods on two large benchmark datasets by a significant margin without degrading the inference speed dramatically.

Luis A. Leiva, Yunfei Xue, Avya Bansal et al.

For graphical user interface (UI) design, it is important to understand what attracts visual attention. While previous work on saliency has focused on desktop and web-based UIs, mobile app UIs differ from these in several respects. We present findings from a controlled study with 30 participants and 193 mobile UIs. The results speak to a role of expectations in guiding where users look at. Strong bias toward the top-left corner of the display, text, and images was evident, while bottom-up features such as color or size affected saliency less. Classic, parameter-free saliency models showed a weak fit with the data, and data-driven models improved significantly when trained specifically on this dataset (e.g., NSS rose from 0.66 to 0.84). We also release the first annotated dataset for investigating visual saliency in mobile UIs.

Nannan Zou, Honglei Zhang, Francesco Cricri et al.

In this paper we present an end-to-end meta-learned system for image compression. Traditional machine learning based approaches to image compression train one or more neural network for generalization performance. However, at inference time, the encoder or the latent tensor output by the encoder can be optimized for each test image. This optimization can be regarded as a form of adaptation or benevolent overfitting to the input content. In order to reduce the gap between training and inference conditions, we propose a new training paradigm for learned image compression, which is based on meta-learning. In a first phase, the neural networks are trained normally. In a second phase, the Model-Agnostic Meta-learning approach is adapted to the specific case of image compression, where the inner-loop performs latent tensor overfitting, and the outer loop updates both encoder and decoder neural networks based on the overfitting performance. Furthermore, after meta-learning, we propose to overfit and cluster the bias terms of the decoder on training image patches, so that at inference time the optimal content-specific bias terms can be selected at encoder-side. Finally, we propose a new probability model for lossless compression, which combines concepts from both multi-scale and super-resolution probability model approaches. We show the benefits of all our proposed ideas via carefully designed experiments.

Sen He, Wentong Liao, Hamed R. Tavakoli et al.

Automatic captioning of images is a task that combines the challenges of image analysis and text generation. One important aspect in captioning is the notion of attention: How to decide what to describe and in which order. Inspired by the successes in text analysis and translation, previous work have proposed the \textit{transformer} architecture for image captioning. However, the structure between the \textit{semantic units} in images (usually the detected regions from object detection model) and sentences (each single word) is different. Limited work has been done to adapt the transformer's internal architecture to images. In this work, we introduce the \textbf{\textit{image transformer}}, which consists of a modified encoding transformer and an implicit decoding transformer, motivated by the relative spatial relationship between image regions. Our design widen the original transformer layer's inner architecture to adapt to the structure of images. With only regions feature as inputs, our model achieves new state-of-the-art performance on both MSCOCO offline and online testing benchmarks.

Nannan Zou, Honglei Zhang, Francesco Cricri et al.

One of the core components of conventional (i.e., non-learned) video codecs consists of predicting a frame from a previously-decoded frame, by leveraging temporal correlations. In this paper, we propose an end-to-end learned system for compressing video frames. Instead of relying on pixel-space motion (as with optical flow), our system learns deep embeddings of frames and encodes their difference in latent space. At decoder-side, an attention mechanism is designed to attend to the latent space of frames to decide how different parts of the previous and current frame are combined to form the final predicted current frame. Spatially-varying channel allocation is achieved by using importance masks acting on the feature-channels. The model is trained to reduce the bitrate by minimizing a loss on importance maps and a loss on the probability output by a context model for arithmetic coding. In our experiments, we show that the proposed system achieves high compression rates and high objective visual quality as measured by MS-SSIM and PSNR. Furthermore, we provide ablation studies where we highlight the contribution of different components.

Alexandre Bruckert, Hamed R. Tavakoli, Zhi Liu et al.

Recent advances in deep learning have pushed the performances of visual saliency models way further than it has ever been. Numerous models in the literature present new ways to design neural networks, to arrange gaze pattern data, or to extract as much high and low-level image features as possible in order to create the best saliency representation. However, one key part of a typical deep learning model is often neglected: the choice of the loss function. In this work, we explore some of the most popular loss functions that are used in deep saliency models. We demonstrate that on a fixed network architecture, modifying the loss function can significantly improve (or depreciate) the results, hence emphasizing the importance of the choice of the loss function when designing a model. We also introduce new loss functions that have never been used for saliency prediction to our knowledge. And finally, we show that a linear combination of several well-chosen loss functions leads to significant improvements in performances on different datasets as well as on a different network architecture, hence demonstrating the robustness of a combined metric.

Zakaria Laskar, Iaroslav Melekhov, Hamed R. Tavakoli et al.

This paper addresses the problem of determining dense pixel correspondences between two images and its application to geometric correspondence verification in image retrieval. The main contribution is a geometric correspondence verification approach for re-ranking a shortlist of retrieved database images based on their dense pair-wise matching with the query image at a pixel level. We determine a set of cyclically consistent dense pixel matches between the pair of images and evaluate local similarity of matched pixels using neural network based image descriptors. Final re-ranking is based on a novel similarity function, which fuses the local similarity metric with a global similarity metric and a geometric consistency measure computed for the matched pixels. For dense matching our approach utilizes a modified version of a recently proposed dense geometric correspondence network (DGC-Net), which we also improve by optimizing the architecture. The proposed model and similarity metric compare favourably to the state-of-the-art image retrieval methods. In addition, we apply our method to the problem of long-term visual localization demonstrating promising results and generalization across datasets.

Hamed R. Tavakoli, Esa Rahtu, Juho Kannala et al.

This paper digs deeper into factors that influence egocentric gaze. Instead of training deep models for this purpose in a blind manner, we propose to inspect factors that contribute to gaze guidance during daily tasks. Bottom-up saliency and optical flow are assessed versus strong spatial prior baselines. Task-specific cues such as vanishing point, manipulation point, and hand regions are analyzed as representatives of top-down information. We also look into the contribution of these factors by investigating a simple recurrent neural model for ego-centric gaze prediction. First, deep features are extracted for all input video frames. Then, a gated recurrent unit is employed to integrate information over time and to predict the next fixation. We also propose an integrated model that combines the recurrent model with several top-down and bottom-up cues. Extensive experiments over multiple datasets reveal that (1) spatial biases are strong in egocentric videos, (2) bottom-up saliency models perform poorly in predicting gaze and underperform spatial biases, (3) deep features perform better compared to traditional features, (4) as opposed to hand regions, the manipulation point is a strong influential cue for gaze prediction, (5) combining the proposed recurrent model with bottom-up cues, vanishing points and, in particular, manipulation point results in the best gaze prediction accuracy over egocentric videos, (6) the knowledge transfer works best for cases where the tasks or sequences are similar, and (7) task and activity recognition can benefit from gaze prediction. Our findings suggest that (1) there should be more emphasis on hand-object interaction and (2) the egocentric vision community should consider larger datasets including diverse stimuli and more subjects.

Sen He, Hamed R. Tavakoli, Ali Borji et al.

Recently, data-driven deep saliency models have achieved high performance and have outperformed classical saliency models, as demonstrated by results on datasets such as the MIT300 and SALICON. Yet, there remains a large gap between the performance of these models and the inter-human baseline. Some outstanding questions include what have these models learned, how and where they fail, and how they can be improved. This article attempts to answer these questions by analyzing the representations learned by individual neurons located at the intermediate layers of deep saliency models. To this end, we follow the steps of existing deep saliency models, that is borrowing a pre-trained model of object recognition to encode the visual features and learning a decoder to infer the saliency. We consider two cases when the encoder is used as a fixed feature extractor and when it is fine-tuned, and compare the inner representations of the network. To study how the learned representations depend on the task, we fine-tune the same network using the same image set but for two different tasks: saliency prediction versus scene classification. Our analyses reveal that: 1) some visual regions (e.g. head, text, symbol, vehicle) are already encoded within various layers of the network pre-trained for object recognition, 2) using modern datasets, we find that fine-tuning pre-trained models for saliency prediction makes them favor some categories (e.g. head) over some others (e.g. text), 3) although deep models of saliency outperform classical models on natural images, the converse is true for synthetic stimuli (e.g. pop-out search arrays), an evidence of significant difference between human and data-driven saliency models, and 4) we confirm that, after-fine tuning, the change in inner-representations is mostly due to the task and not the domain shift in the data.

Zakaria Laskar, Hamed R. Tavakoli, Juho Kannala

In this paper we address the problem of establishing correspondences between different instances of the same object. The problem is posed as finding the geometric transformation that aligns a given image pair. We use a convolutional neural network (CNN) to directly regress the parameters of the transformation model. The alignment problem is defined in the setting where an unordered set of semantic key-points per image are available, but, without the correspondence information. To this end we propose a novel loss function based on cyclic consistency that solves this 2D point set registration problem by inferring the optimal geometric transformation model parameters. We train and test our approach on a standard benchmark dataset Proposal-Flow (PF-PASCAL)\cite{proposal_flow}. The proposed approach achieves state-of-the-art results demonstrating the effectiveness of the method. In addition, we show our approach further benefits from additional training samples in PF-PASCAL generated by using category level information.

Ali Borji, Hamed R. Tavakoli, Zoya Bylinskii

In this review, we examine the recent progress in saliency prediction and proposed several avenues for future research. In spite of tremendous efforts and huge progress, there is still room for improvement in terms finer-grained analysis of deep saliency models, evaluation measures, datasets, annotation methods, cognitive studies, and new applications. This chapter will appear in Encyclopedia of Computational Neuroscience.

Hamed R. Tavakoli, Fawad Ahmed, Ali Borji et al.

This paper revisits visual saliency prediction by evaluating the recent advancements in this field such as crowd-sourced mouse tracking-based databases and contextual annotations. We pursue a critical and quantitative approach towards some of the new challenges including the quality of mouse tracking versus eye tracking for model training and evaluation. We extend quantitative evaluation of models in order to incorporate contextual information by proposing an evaluation methodology that allows accounting for contextual factors such as text, faces, and object attributes. The proposed contextual evaluation scheme facilitates detailed analysis of models and helps identify their pros and cons. Through several experiments, we find that (1) mouse tracking data has lower inter-participant visual congruency and higher dispersion, compared to the eye tracking data, (2) mouse tracking data does not totally agree with eye tracking in general and in terms of different contextual regions in specific, and (3) mouse tracking data leads to acceptable results in training current existing models, and (4) mouse tracking data is less reliable for model selection and evaluation. The contextual evaluation also reveals that, among the studied models, there is no single model that performs best on all the tested annotations.

Hamed R. Tavakoli, Rakshith Shetty, Ali Borji et al.

To bridge the gap between humans and machines in image understanding and describing, we need further insight into how people describe a perceived scene. In this paper, we study the agreement between bottom-up saliency-based visual attention and object referrals in scene description constructs. We investigate the properties of human-written descriptions and machine-generated ones. We then propose a saliency-boosted image captioning model in order to investigate benefits from low-level cues in language models. We learn that (1) humans mention more salient objects earlier than less salient ones in their descriptions, (2) the better a captioning model performs, the better attention agreement it has with human descriptions, (3) the proposed saliency-boosted model, compared to its baseline form, does not improve significantly on the MS COCO database, indicating explicit bottom-up boosting does not help when the task is well learnt and tuned on a data, (4) a better generalization is, however, observed for the saliency-boosted model on unseen data.

Hamed R. Tavakoli, Jorma Laaksonen

This manuscript introduces the problem of prominent object detection and recognition inspired by the fact that human seems to priorities perception of scene elements. The problem deals with finding the most important region of interest, segmenting the relevant item/object in that area, and assigning it an object class label. In other words, we are solving the three problems of saliency modeling, saliency detection, and object recognition under one umbrella. The motivation behind such a problem formulation is (1) the benefits to the knowledge representation-based vision pipelines, and (2) the potential improvements in emulating bio-inspired vision systems by solving these three problems together. We are foreseeing extending this problem formulation to fully semantically segmented scenes with instance object priority for high-level inferences in various applications including assistive vision. Along with a new problem definition, we also propose a method to achieve such a task. The proposed model predicts the most important area in the image, segments the associated objects, and labels them. The proposed problem and method are evaluated against human fixations, annotated segmentation masks, and object class categories. We define a chance level for each of the evaluation criterion to compare the proposed algorithm with. Despite the good performance of the proposed baseline, the overall evaluations indicate that the problem of prominent object detection and recognition is a challenging task that is still worth investigating further.

Hamed R. Tavakoli, Jorma Laaksonen, Esa Rahtu

This paper revisits recognition of natural image pleasantness by employing deep convolutional neural networks and affordable eye trackers. There exist several approaches to recognize image pleasantness: (1) computer vision, and (2) psychophysical signals. For natural images, computer vision approaches have not been as successful as for abstract paintings and is lagging behind the psychophysical signals like eye movements. Despite better results, the scalability of eye movements is adversely affected by the sensor cost. While the introduction of affordable sensors have helped the scalability issue by making the sensors more accessible, the application of such sensors in a loosely controlled human-computer interaction setup is not yet studied for affective image tagging. On the other hand, deep convolutional neural networks have boosted the performance of vision-based techniques significantly in recent years. To investigate the current status in regard to affective image tagging, we (1) introduce a new eye movement dataset using an affordable eye tracker, (2) study the use of deep neural networks for pleasantness recognition, (3) investigate the gap between deep features and eye movements. To meet these ends, we record eye movements in a less controlled setup, akin to daily human-computer interaction. We assess features from eye movements, visual features, and their combination. Our results show that (1) recognizing natural image pleasantness from eye movement under less restricted setup is difficult and previously used techniques are prone to fail, and (2) visual class categories are strong cues for predicting pleasantness, due to their correlation with emotions, necessitating careful study of this phenomenon. This latter finding is alerting as some deep learning approaches may fit to the class category bias.