pith. sign in

arxiv: 2604.10885 · v1 · submitted 2026-04-13 · 💻 cs.CV · cs.AI· cs.GR

Product Review Based on Optimized Facial Expression Detection

Vikrant Chaugule , Abhishek D , Aadheeshwar Vijayakumar , Pravin Bhaskar Ramteke , Shashidhar G. Koolagudi This is my paper

Pith reviewed 2026-05-10 15:36 UTC · model grok-4.3

classification 💻 cs.CV cs.AIcs.GR
keywords algorithmexpressionfacialdetectionharrisproductreviewtime
0
0 comments X

The pith

A modified Harris algorithm extracts facial features faster for product review via customer expressions.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper aims to use facial expression analysis to determine customer acceptance of products in retail environments. It introduces a modified version of the Harris corner detection algorithm to identify key facial points more efficiently. By lowering the time complexity of feature extraction, the method enables quicker processing suitable for real-time applications. A comparison with existing algorithms demonstrates the speed improvement while claiming near accuracy for the task at hand.

Core claim

The central discovery is that modifying the Harris algorithm reduces the time complexity associated with detecting corner points in images, allowing facial expression detection to be performed significantly faster and with nearly the same accuracy as the original method, making it viable for reviewing product acceptance based on customer facial reactions in supermarkets.

What carries the argument

The modified Harris algorithm, which optimizes the detection of corner points by reducing computational steps in feature extraction for facial images.

Load-bearing premise

That a modification to the Harris corner detector can substantially reduce time complexity while retaining enough accuracy to reliably identify facial expressions for product acceptance judgments.

What would settle it

Running the modified algorithm on a standard facial expression dataset and comparing its accuracy and processing time to both the original Harris method and modern expression recognition techniques.

Figures

Figures reproduced from arXiv: 2604.10885 by Aadheeshwar Vijayakumar, Abhishek D, Pravin Bhaskar Ramteke, Shashidhar G. Koolagudi, Vikrant Chaugule.

Figure 1
Figure 1. Figure 1: Haar like feature high probability that face is present and is shown in figure 2. Mouth detection-Mouth is detected in same manner as the face is detected. Eye detection-Eye is detected in same manner as the face is detected. B. EMOTION DETECTION Emotion recognition is the process of identifying hu￾man emotion, most typically from facial expressions. This is both something that humans do automatically but … view at source ↗
Figure 2
Figure 2. Figure 2: Cascade of feature Classifiers [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Graph showing high sigma values require significantly more [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: flow chart of the proposed approach take a lot of time to get a feedback of their product. Usually by the time they get a public opinion of their product the brand image of the company goes down. Also for new products there is a delay involved in finding out the consumers perception of the product launched. The proposed approach helps companies to take up quick action to rectify the shortcomings of the pro… view at source ↗
Figure 5
Figure 5. Figure 5: Corner Detection Theory . For the eyes, we iterate through all feature points y(j) for each x(i) and then maximum and minimum values are appropriately calculated. Using these values, the curious ratio for eyes is calculated. The same procedure followed for feature points of the mouth and the curious ratio for mouth is calculated. Both of these values help in deciding the curiousness level of the person. Al… view at source ↗
Figure 7
Figure 7. Figure 7: Comparison on sad face V. Evaluation of proposed approach (Survey) The approach for product review using emotion de￾tection has been done properly for the first time and hence comparative study is not required. The comparison for feature extraction algorithm over the previous one is required as its a new approach suitable for product review and known for its speed. These following results are verified by r… view at source ↗
Figure 6
Figure 6. Figure 6: Comparison on happy face [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: Table showing time required for processing [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Graph showing time required for processing [PITH_FULL_IMAGE:figures/full_fig_p007_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Table showing comparison of accuracy over previous algorithm [PITH_FULL_IMAGE:figures/full_fig_p008_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Graph showing comparison of accuracy over previous algorithm [PITH_FULL_IMAGE:figures/full_fig_p008_11.png] view at source ↗
read the original abstract

This paper proposes a method to review public acceptance of products based on their brand by analyzing the facial expression of the customer intending to buy the product from a supermarket or hypermarket. In such cases, facial expression recognition plays a significant role in product review. Here, facial expression detection is performed by extracting feature points using a modified Harris algorithm. The modified Harris algorithm reduced the time complexity of the existing feature extraction Harris Algorithm. A comparison of time complexities of existing algorithms is done with proposed algorithm. The algorithm proved to be significantly faster and nearly accurate for the needed application by reducing the time complexity for corner points detection.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 0 minor

Summary. The paper proposes a method for reviewing product acceptance in supermarkets by analyzing customers' facial expressions using a modified Harris algorithm for feature extraction. It claims this modification reduces time complexity compared to the standard Harris corner detector, with a comparison of time complexities showing the proposed algorithm to be significantly faster and nearly accurate for the application.

Significance. If the claims of reduced time complexity and sufficient accuracy are substantiated, the method could offer a practical tool for real-time market research in retail settings by enabling efficient facial expression recognition without heavy computational resources. However, the current lack of supporting evidence diminishes its immediate significance.

major comments (3)
  1. Abstract: The central claim that 'the modified Harris algorithm reduced the time complexity of the existing feature extraction Harris Algorithm' provides no description of the modification, no O-notation analysis, and no derivation or implementation details.
  2. Abstract: The assertion that 'a comparison of time complexities of existing algorithms is done with proposed algorithm' is unsupported, as the manuscript contains no tables, figures, numerical runtime values, or methodological comparison.
  3. Abstract: No quantitative results (accuracy rates, error metrics, time measurements), dataset details, or validation on facial expression recognition benchmarks are reported to support the claim that the algorithm is 'significantly faster and nearly accurate'.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the thorough review and valuable feedback. We agree that the submitted manuscript, particularly the abstract, lacks essential details on the algorithm modification, complexity analysis, comparisons, and empirical results. We will prepare a major revision that expands these sections with the required descriptions, derivations, tables, and quantitative evaluations to substantiate the claims.

read point-by-point responses

  1. Referee: Abstract: The central claim that 'the modified Harris algorithm reduced the time complexity of the existing feature extraction Harris Algorithm' provides no description of the modification, no O-notation analysis, and no derivation or implementation details.

    Authors: We acknowledge this point. The abstract was kept brief, but the full manuscript does not provide the requested details. In the revised version we will add a dedicated subsection describing the specific modifications to the Harris corner detector (e.g., the optimized corner response function and reduced search window), derive the improved time complexity using O-notation, and include the step-by-step implementation details. revision: yes

  2. Referee: Abstract: The assertion that 'a comparison of time complexities of existing algorithms is done with proposed algorithm' is unsupported, as the manuscript contains no tables, figures, numerical runtime values, or methodological comparison.

    Authors: The referee correctly notes the absence of supporting material. We will insert a new comparison section containing a table of asymptotic complexities for standard Harris, FAST, SIFT and the proposed variant, together with any available empirical runtime measurements obtained from our implementation on the same hardware. revision: yes

  3. Referee: Abstract: No quantitative results (accuracy rates, error metrics, time measurements), dataset details, or validation on facial expression recognition benchmarks are reported to support the claim that the algorithm is 'significantly faster and nearly accurate'.

    Authors: We agree that quantitative validation is missing. The revision will include an experimental results section reporting accuracy, precision, recall and F1 scores on standard facial expression datasets (e.g., CK+, JAFFE), execution-time measurements, and direct comparison against baseline detectors to demonstrate that the method remains sufficiently accurate for the retail application while being faster. revision: yes

Circularity Check

0 steps flagged

No derivation chain or mathematical structure present; claims are bare assertions

full rationale

The paper asserts that a modified Harris algorithm reduces time complexity for corner-point detection and is significantly faster while nearly accurate, with a comparison to existing algorithms mentioned but not shown. No equations, no description of the modification, no O-notation analysis, no runtime measurements, and no self-citations appear in the text. Because there is no derivation chain, fitted parameters, or load-bearing premises to inspect, no step reduces to its own inputs by construction. The finding is therefore a non-finding of circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper provides no mathematical derivation, so the ledger is nearly empty. The only implicit assumption is that facial expressions can be reliably mapped to product acceptance via corner features.

axioms (1)
  • domain assumption Facial expressions captured in retail settings reliably indicate product acceptance
    Stated in the abstract as the basis for using expression detection to produce product reviews.

pith-pipeline@v0.9.0 · 5418 in / 1203 out tokens · 50349 ms · 2026-05-10T15:36:56.212900+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

22 extracted references · 22 canonical work pages · 1 internal anchor

  1. [1]

    Facial emotion recognition using multi-modal information

    Liyanage C De Silva, Tsutomu Miyasato, and Ry- ohei Nakatsu. Facial emotion recognition using multi-modal information. InInformation, Com- munications and Signal Processing, 1997. ICICS., Proceedings of 1997 International Conference on, volume 1, pages 397–401. IEEE, 1997

    work page 1997

  2. [2]

    Facial ex- pression recognition in adolescents with mood and anxiety disorders.American Journal of Psychiatry, 2003

    Erin B McClure, Kayla Pope, Andrea J Hoberman, Daniel S Pine, and Ellen Leibenluft. Facial ex- pression recognition in adolescents with mood and anxiety disorders.American Journal of Psychiatry, 2003

    work page 2003

  3. [3]

    Adaptive harris corner detec- tion algorithm.Computer Engineering, 10(5):212– 215, 2008

    Wan-jin Zhao, Sheng-rong Gong, Chun-ping Liu, and Xiang-jun SHEN. Adaptive harris corner detec- tion algorithm.Computer Engineering, 10(5):212– 215, 2008

    work page 2008

  4. [4]

    A combined corner and edge detector

    Chris Harris and Mike Stephens. A combined corner and edge detector. InAlvey vision conference, volume 15, page 50. Citeseer, 1988

    work page 1988

  5. [5]

    Matlab and octave functions for computer vision and image processing.Online: http://www

    Peter D Kovesi. Matlab and octave functions for computer vision and image processing.Online: http://www. csse. uwa. edu. au/˜ pk/Research/Mat- labFns/# match, 2000

    work page 2000

  6. [6]

    Generative ai for video trailer synthesis: From extractive heuristics to autoregressive creativity.Authorea Preprints, 2026

    Abhishek Dharmaratnakar, Srivaths Ranganathan, Debanshu Das, and Anushree Sinha. Generative ai for video trailer synthesis: From extractive heuristics to autoregressive creativity.Authorea Preprints, 2026

    work page 2026

  7. [7]

    Dsl approach for development of gaming applications

    Aadheeshwar Vijayakumar, D Abhishek, and K Chandrasekaran. Dsl approach for development of gaming applications. InInformation Systems Design and Intelligent Applications: Proceedings of Third International Conference INDIA 2016, Volume 1, pages 199–211. Springer India New Delhi, 2016

    work page 2016

  8. [8]

    Good features to track

    Jianbo Shi and Carlo Tomasi. Good features to track. InComputer Vision and Pattern Recognition,

    work page

  9. [9]

    IEEE, 1994

    Proceedings CVPR’94., 1994 IEEE Computer Society Conference on, pages 593–600. IEEE, 1994

    work page 1994

  10. [10]

    Robust real-time face detection.International journal of computer vision, 57(2):137–154, 2004

    Paul Viola and Michael J Jones. Robust real-time face detection.International journal of computer vision, 57(2):137–154, 2004

    work page 2004

  11. [11]

    An improved corner detec- tion method based on harris [j].Computer Technol- ogy and Development, 5:036, 2009

    Yan-ming Mao, Mei-hui LAN, Yun-qiong W ANG, and Qiao-sheng FENG. An improved corner detec- tion method based on harris [j].Computer Technol- ogy and Development, 5:036, 2009

    work page 2009

  12. [12]

    The evolutionary psychology of facial beauty.Annu

    Gillian Rhodes. The evolutionary psychology of facial beauty.Annu. Rev. Psychol., 57:199–226, 2006

    work page 2006

  13. [13]

    The development of facial emotion recognition: The role of configural information.Journal of experimental child psychology, 97(1):14–27, 2007

    Karine Durand, Mathieu Gallay, Alix Seigneuric, Fabrice Robichon, and Jean-Yves Baudouin. The development of facial emotion recognition: The role of configural information.Journal of experimental child psychology, 97(1):14–27, 2007

    work page 2007

  14. [14]

    Vlfeat: An open and portable library of computer vision algo- rithms

    Andrea Vedaldi and Brian Fulkerson. Vlfeat: An open and portable library of computer vision algo- rithms. InProceedings of the 18th ACM interna- tional conference on Multimedia, pages 1469–1472. ACM, 2010

    work page 2010

  15. [15]

    Face de- tection and smile detection

    Yu-Hao Huang and Chiou-Shann Fuh. Face de- tection and smile detection. InProceedings of IPPR Conference on Computer Vision, Graphics and Image Porcessing, Shitou, Taiwan, A5-6, page 108, 2009

    work page 2009

  16. [16]

    Fast corner detection.Image and vision computing, 16(2):75– 87, 1998

    Miroslav Trajkovi ´c and Mark Hedley. Fast corner detection.Image and vision computing, 16(2):75– 87, 1998

    work page 1998

  17. [17]

    Real-time corner detection algorithm for motion estimation.Image and Vision Computing, 13(9):695–703, 1995

    Han Wang and Michael Brady. Real-time corner detection algorithm for motion estimation.Image and Vision Computing, 13(9):695–703, 1995

    work page 1995

  18. [18]

    A review on smiley face recognition

    Mridul Dixit, Piyush Rai, and Sanjay Silakari. A review on smiley face recognition. InCommuni- cation Systems and Network Technologies (CSNT), 2013 International Conference on, pages 137–139. IEEE, 2013

    work page 2013

  19. [19]

    ”Multi-Agent Video Recommenders: Evo- lution, Patterns, and Open Challenges.” arXiv preprint arXiv:2604.02211 (2026)

    Srivaths Ranganathan, Abhishek Dharmaratnakar, Anushree Sinha, and Debanshu Das. Multi-agent video recommenders: Evolution, patterns, and open challenges.arXiv preprint arXiv:2604.02211, 2026

    work page arXiv 2026

  20. [20]

    Beyond Fluency: Toward Reliable Trajectories in Agentic IR

    Anushree Sinha, Srivaths Ranganathan, Debanshu Das, and Abhishek Dharmaratnakar. Beyond flu- ency: Toward reliable trajectories in agentic ir.arXiv preprint arXiv:2604.04269, 2026

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  21. [21]

    A semantic approach to text steganography in sanskrit using numerical encoding

    K Vaishakh, A Pravalika, DV Abhishek, NP Meghana, and Gaurav Prasad. A semantic approach to text steganography in sanskrit using numerical encoding. InRecent Findings in Intelligent Computing Techniques: Proceedings of the 5th ICACNI 2017, Volume 1, pages 181–192. Springer Singapore Singapore, 2018

    work page 2017

  22. [22]

    Factuality and hallucinations in large language models: A compre- hensive survey.Authorea Preprints, 2026

    Srivaths Ranganathan, Abhishek Dharmaratnakar, Anushree Sinha, and Debanshu Das. Factuality and hallucinations in large language models: A compre- hensive survey.Authorea Preprints, 2026

    work page 2026